university of groningen potato starch stabilized synthetic ......potato starch stabilized synthetic...
TRANSCRIPT
University of Groningen
Potato starch stabilized synthetic latexesTerpstra, Karsjen
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.
Document VersionPublisher's PDF, also known as Version of record
Publication date:2015
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):Terpstra, K. (2015). Potato starch stabilized synthetic latexes. [S.l.]: [S.n.].
CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.
Download date: 12-10-2020
Potato starch stabilized synthetic latexes
Ronald Terpstra
Potato starch stabilized synthetic latexes
Proefschrift
ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen
op gezag van de rector magnificus prof. dr. E. Sterken
en volgens besluit van het College voor Promoties.
De openbare verdediging zal plaatsvinden op
vrijdag 20 maart 2015 om 14.30 uur
door
Karsjen Ronald Terpstra
geboren op 28 februari 1970 te Zuidlaren
This investigation was financially supported by AVEBE U.A., Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
Cover design: Henk F. Terpstra (De opspringende natuur en het inspringend onderzoek)Layout: Claudia González Arévalo ([email protected])Printing: GrafiMedia
ISBN: 978-90-367-7640-0ISBN (Electronic version): 978-90-367-7639-4
Potato starch stabilized synthetic latexes
Proefschrift
ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen
op gezag van de rector magnificus prof. dr. E. Sterken
en volgens besluit van het College voor Promoties.
De openbare verdediging zal plaatsvinden op
vrijdag 20 maart 2015 om 14.30 uur
door
Karsjen Ronald Terpstra
geboren op 28 februari 1970 te Zuidlaren
This investigation was financially supported by AVEBE U.A., Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
Cover design: Henk F. Terpstra (De opspringende natuur en het inspringend onderzoek)Layout: Claudia González Arévalo ([email protected])Printing: GrafiMedia
ISBN: 978-90-367-7640-0ISBN (Electronic version): 978-90-367-7639-4
“A change can be as good as a rest “
Promotores Prof. dr. F. Picchioni Prof. dr. ir. H.J. Heeres
Beoordelingscommissie Prof. dr. K.U. Loos Prof. dr. A.A. Broekhuis Prof. dr. L. Moscicki
“A change can be as good as a rest “
Promotores Prof. dr. F. Picchioni Prof. dr. ir. H.J. Heeres
Beoordelingscommissie Prof. dr. K.U. Loos Prof. dr. A.A. Broekhuis Prof. dr. L. Moscicki
Chapter 5 Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and waxy potato starch
85
Introduction 86
Experimental 88
Results 90
Conclusions 98
Abbreviations 98
References 100
Chapter 6 The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
101
Introduction 102
Experimental 104
Results 105
Conclusions 110
Acknowledgements 110
Abbreviations 111
References 112
Chapter 7 Technological assessment 113Starch as protective colloid 114
Free radical reactor configuration and polymerization procedure
114
Final product 115
References 118
Appendices Summary 121
Samenvatting 127
Dankwoord 133
Publications 139
Table of contentsChapter 1 Introduction 9
Latex 10
Green chemistry and engineering 13
Starch 16
Modified starch in polyvinyl acetate based latexes 20
Scope of this thesis 22
Abbreviations 23
References 24
Chapter 2 Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
27
Introduction 28
Experimental 30
Results 34
Conclusions 44
Acknowledgements 45
Abbreviations 45
References 46
Appendix 49
Chapter 3 A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
51
Introduction 52
Experimental 55
Results 59
Conclusions 64
Acknowledgements 65
Abbreviations 65
References 66
Appendix 68
Chapter 4 Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics
69
Introduction 70
Experimental 72
Results 74
Conclusions 81
Abbreviations 82
References 83
Chapter 5 Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and waxy potato starch
85
Introduction 86
Experimental 88
Results 90
Conclusions 98
Abbreviations 98
References 100
Chapter 6 The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
101
Introduction 102
Experimental 104
Results 105
Conclusions 110
Acknowledgements 110
Abbreviations 111
References 112
Chapter 7 Technological assessment 113Starch as protective colloid 114
Free radical reactor configuration and polymerization procedure
114
Final product 115
References 118
Appendices Summary 121
Samenvatting 127
Dankwoord 133
Publications 139
Table of contentsChapter 1 Introduction 9
Latex 10
Green chemistry and engineering 13
Starch 16
Modified starch in polyvinyl acetate based latexes 20
Scope of this thesis 22
Abbreviations 23
References 24
Chapter 2 Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
27
Introduction 28
Experimental 30
Results 34
Conclusions 44
Acknowledgements 45
Abbreviations 45
References 46
Appendix 49
Chapter 3 A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
51
Introduction 52
Experimental 55
Results 59
Conclusions 64
Acknowledgements 65
Abbreviations 65
References 66
Appendix 68
Chapter 4 Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics
69
Introduction 70
Experimental 72
Results 74
Conclusions 81
Abbreviations 82
References 83
CHAPTER 1Introduction
CHAPTER 1Introduction
Introduction
10 11
Table 2:The different types of heterogeneous polymerization systems [1].
TypeParticle radius (nm)
Droplet size(μm)
Initiator Continuous phase
Discrete phase (particles)
* Emulsion 50 - 300 ~1 - 10 Water soluble Water
Initially absent, monomer swollen polymer particles
* Precipitation 50 - 300Water soluble monomer
Water soluble Water
Emulsion but monomer does not swell polymer
* Suspension ≥ 1000 1 – 10 Oil soluble Water
Monomer + formed polymer in pre-existing droplets
* Dispersion ≥ 1000 - Water soluble
Organic (poor polymer solvent)
Initially absent, monomer-swollen polymer particles.
* Microemulsion 10 - 30 ~0.01 Water soluble Water
Monomer, co-surfactant + formed polymer
* Inverse emulsion 100 - 1000 1 - 10Water or oil soluble
Oil Monomer, water+ formed polymer
* mini-emulsion 30 - 100 ~0.03 Water soluble Water
Monomer, co-surfactant + formed polymer
Emulsion polymerization requires a proper distribution of the monomer in water during processing and this is usually achieved by adding a component with detergent or emulsifier characteristics. The actual synthesis is controlled by heterogeneously generated free radicals and the polymerization can be divided in three characteristical stages as shown in Figure 1.
Figure 1: A typical rate of polymerization as a function of the monomer conversion in combination with a schematic representation of the micelle nucleation model [3].
LatexA latex is a stable dispersion of small polymer particles in water and can originate from either a natural or synthetic source. The synthetic varieties can be obtained by free radical polymerization of monomers containing a vinyl (i.e ethenyl according to the formal IUPAC rule) double bond. The structures of common commercial available monomers are given in Table 1 together with an indication of their area of application [1].
Table 1:Typical monomers used in common commercial emulsion polymerizations [1].Monomer Structure Application area
* Styrene Synthetic rubbers; Paper coating
* Butadiene Synthetic rubbers; Paper coating
* Tetrafluoroethylene Poly(tetrafluoroethylene) (e.g Teflon)Fluoropolymers (e.g. Viton)
* Vinyl acetate Adhesive; paint
* Methyl methacrylate Surface coating
* Acrylic acid Paint
* Itaconic acid Paint
* Chloroprene Neoprene rubber
* Butyl acrylate Surface coating
* Butyl methacrylate Surface coating
* Methyl acrylate Surface coating; Adhesive
* Vinyl chloride Water, sewage and electrical tubing
The differences between the monomers (e.g. reactivity and water solubility) do not only have an impact on process characteristics but also influence the properties of the final polymer (e.g. glass transition temperature and hydrophobicity). Even combinations of monomers can be used to meet the desired molecular characteristics [2]. Table 2 shows a number of methodologies in which these latexes can be produced. Emulsion polymerization is the most used technique for preparing latexes used as coatings and adhesive [1].
F
FF
F
O
O
O
OHO
O
HOOCCOOH
Cl
O
O
O
OO
O
ClR
R R
Legend:
- M = Monomer; R = Radical; P = Polymer.
- Dashed circle: Hydrophilic part of polymer and detergent.
Stage I: Nucleation
Stage II: Growth Stage III: Consumption
Polym
eriza
tion r
ate
Monomer conversion (%)
MMP MPMMP MPM
MM MM
M M M M M M M M M
MPMP MPPPPM PPPMPP MPMP
0 100
Introduction
10 11
Table 2:The different types of heterogeneous polymerization systems [1].
TypeParticle radius (nm)
Droplet size(μm)
Initiator Continuous phase
Discrete phase (particles)
* Emulsion 50 - 300 ~1 - 10 Water soluble Water
Initially absent, monomer swollen polymer particles
* Precipitation 50 - 300Water soluble monomer
Water soluble Water
Emulsion but monomer does not swell polymer
* Suspension ≥ 1000 1 – 10 Oil soluble Water
Monomer + formed polymer in pre-existing droplets
* Dispersion ≥ 1000 - Water soluble
Organic (poor polymer solvent)
Initially absent, monomer-swollen polymer particles.
* Microemulsion 10 - 30 ~0.01 Water soluble Water
Monomer, co-surfactant + formed polymer
* Inverse emulsion 100 - 1000 1 - 10Water or oil soluble
Oil Monomer, water+ formed polymer
* mini-emulsion 30 - 100 ~0.03 Water soluble Water
Monomer, co-surfactant + formed polymer
Emulsion polymerization requires a proper distribution of the monomer in water during processing and this is usually achieved by adding a component with detergent or emulsifier characteristics. The actual synthesis is controlled by heterogeneously generated free radicals and the polymerization can be divided in three characteristical stages as shown in Figure 1.
Figure 1: A typical rate of polymerization as a function of the monomer conversion in combination with a schematic representation of the micelle nucleation model [3].
LatexA latex is a stable dispersion of small polymer particles in water and can originate from either a natural or synthetic source. The synthetic varieties can be obtained by free radical polymerization of monomers containing a vinyl (i.e ethenyl according to the formal IUPAC rule) double bond. The structures of common commercial available monomers are given in Table 1 together with an indication of their area of application [1].
Table 1:Typical monomers used in common commercial emulsion polymerizations [1].Monomer Structure Application area
* Styrene Synthetic rubbers; Paper coating
* Butadiene Synthetic rubbers; Paper coating
* Tetrafluoroethylene Poly(tetrafluoroethylene) (e.g Teflon)Fluoropolymers (e.g. Viton)
* Vinyl acetate Adhesive; paint
* Methyl methacrylate Surface coating
* Acrylic acid Paint
* Itaconic acid Paint
* Chloroprene Neoprene rubber
* Butyl acrylate Surface coating
* Butyl methacrylate Surface coating
* Methyl acrylate Surface coating; Adhesive
* Vinyl chloride Water, sewage and electrical tubing
The differences between the monomers (e.g. reactivity and water solubility) do not only have an impact on process characteristics but also influence the properties of the final polymer (e.g. glass transition temperature and hydrophobicity). Even combinations of monomers can be used to meet the desired molecular characteristics [2]. Table 2 shows a number of methodologies in which these latexes can be produced. Emulsion polymerization is the most used technique for preparing latexes used as coatings and adhesive [1].
F
FF
F
O
O
O
OHO
O
HOOCCOOH
Cl
O
O
O
OO
O
ClR
R R
Legend:
- M = Monomer; R = Radical; P = Polymer.
- Dashed circle: Hydrophilic part of polymer and detergent.
Stage I: Nucleation
Stage II: Growth Stage III: Consumption
Polym
eriza
tion r
ate
Monomer conversion (%)
MMP MPMMP MPM
MM MM
M M M M M M M M M
MPMP MPPPPM PPPMPP MPMP
0 100
Introduction
12 13
Table 3: World emulsion polymer demand [5,6].
YearsDemand (metric tons) Growth (%)
2006 2011 2016 2021 06-11 11-16* World demand (thousand metric tons) 9110 10330 13250 16450 2.5 5.1
By marketCoating 5370 6055 7890 9830 2.4 5.4Adhesives 2235 2490 3175 3955 2.2 5.0Other markets 1505 1785 2185 2665 3.5 4.1
By region:North America 2757 2460 3010 3400 -2.3 4.1
United States 2380 2105 2575 2900 -2.4 4.1Canada & Mexico 377 355 435 500 -1.2 4.1
Western Europe 2664 2582 2870 3120 -0.6 2.1Asia/Pacific 2748 4078 5760 7860 8.2 7.2
China 1045 2040 3160 4610 14.3 9.1Japan 628 583 635 660 -1.5 1.7Other 1075 1455 1965 2590 6.2 6.2
Other regions 941 1210 1610 2070 5.2 5.9* Dollar / kg 2.23 2.52 2.74 2.98 2.5 1.7* World demand (million dollar) 20350 26060 36350 49000 5.1 6.9
Green chemistry and engineeringThe “natural step, bio-mimicry, cradle to cradle, getting to zero waste, resilience engineering, inherently safer design, ecological design, green chemistry and self-assembly” are just a number of examples of the abundantly available philosophies for making chemical products in a more sustainable and safe way. The number of institutions and individuals converting these philosophies into workable principles and guidelines is equally overwhelming. However, there are considerable overlaps between these principles and guidelines and some of them are not technically oriented. A selection based on their relevance to understand the fundamental idea of chemical engineering promoting sustainable development is very enlightening [7]. The design for the environment (DFE) appears to be an excellent guideline for all stages of sustainable development even if the proposed points of view appear to be of a more general nature [8]. The twelve principles of green chemistry and engineering are ready-to-use guidelines by which the impact of chemical products on the human and environmental health can be reduced significantly [9,10]. Petrochemical derived detergents, emulsifiers and protective colloids are frequently required during polymerization to achieve the desired latex functionality [3]. Replacing these ingredients by renewable alternatives increases the degree of sustainability of these products according the twelve principles of green chemistry (Table 4). Moreover, the improvement is not always restricted to the actual replacement of this part of the formulation because the latexes obtained are in numerous cases blended with petrochemical based binders and other components to achieve optimal performance in the desired application. Unfortunately, replacement of these blending ingredients by their renewable counterparts is frequently limited due to compatibility issues. However, the risk of compatibility problems is reduced considerably if
The dominating process during stage I is the formation and initiation of monomer-filled soap micelles. This stage corresponds with 10-20% of the monomer addition and determines the number of particles formed. This stage is also important for the properties of the latex since it (indirectly) determines the size of particles present in the final product. The main process in stage II is the growth of particles and in stage III is a diminishing monomer concentration the most important variable. Stabilization of the hydrophobic particles is crucial in all three stages. This can be achieved by either electrostatic repulsion of ionic groups at the polymer/water interface or steric stabilization by hydrophilic polymers (protective colloids), which can generate a so-called protective water barrier around the latex particle (Figure 2). Electrostatic stabilization is complementary to steric stabilization and the two are therefore frequently combined to achieve an optimum result [3,4].
Figure 2:A latex particle covered with degraded starch fragments.
The global demand for emulsion polymers (i.e. latexes) is increasing and studies show that this will continue in the near future (Table 3). The Asia/Pacific region has become the leading regional latex market in the last decade due to the very fast growing economies in India and China. This considerable growth is believed to continue in the years to come whilst the markets in North America and Western Europe are believed to increase at a subpar pace. Positive prospects are also expected in the regions Africa and Middle East.The predicted expansion of the latex market is partly based on the world wide desire to use more environmental benign products. The focus in developing nations is on the replacement of organic solvent-based products by their water-based counterparts whilst the industrialized world, due to the fact that waterborne products are already very common, is focusing on preparing them in a more sustainable way.
Protective water barrier
Particle
Protective water barrier
Introduction
12 13
Table 3: World emulsion polymer demand [5,6].
YearsDemand (metric tons) Growth (%)
2006 2011 2016 2021 06-11 11-16* World demand (thousand metric tons) 9110 10330 13250 16450 2.5 5.1
By marketCoating 5370 6055 7890 9830 2.4 5.4Adhesives 2235 2490 3175 3955 2.2 5.0Other markets 1505 1785 2185 2665 3.5 4.1
By region:North America 2757 2460 3010 3400 -2.3 4.1
United States 2380 2105 2575 2900 -2.4 4.1Canada & Mexico 377 355 435 500 -1.2 4.1
Western Europe 2664 2582 2870 3120 -0.6 2.1Asia/Pacific 2748 4078 5760 7860 8.2 7.2
China 1045 2040 3160 4610 14.3 9.1Japan 628 583 635 660 -1.5 1.7Other 1075 1455 1965 2590 6.2 6.2
Other regions 941 1210 1610 2070 5.2 5.9* Dollar / kg 2.23 2.52 2.74 2.98 2.5 1.7* World demand (million dollar) 20350 26060 36350 49000 5.1 6.9
Green chemistry and engineeringThe “natural step, bio-mimicry, cradle to cradle, getting to zero waste, resilience engineering, inherently safer design, ecological design, green chemistry and self-assembly” are just a number of examples of the abundantly available philosophies for making chemical products in a more sustainable and safe way. The number of institutions and individuals converting these philosophies into workable principles and guidelines is equally overwhelming. However, there are considerable overlaps between these principles and guidelines and some of them are not technically oriented. A selection based on their relevance to understand the fundamental idea of chemical engineering promoting sustainable development is very enlightening [7]. The design for the environment (DFE) appears to be an excellent guideline for all stages of sustainable development even if the proposed points of view appear to be of a more general nature [8]. The twelve principles of green chemistry and engineering are ready-to-use guidelines by which the impact of chemical products on the human and environmental health can be reduced significantly [9,10]. Petrochemical derived detergents, emulsifiers and protective colloids are frequently required during polymerization to achieve the desired latex functionality [3]. Replacing these ingredients by renewable alternatives increases the degree of sustainability of these products according the twelve principles of green chemistry (Table 4). Moreover, the improvement is not always restricted to the actual replacement of this part of the formulation because the latexes obtained are in numerous cases blended with petrochemical based binders and other components to achieve optimal performance in the desired application. Unfortunately, replacement of these blending ingredients by their renewable counterparts is frequently limited due to compatibility issues. However, the risk of compatibility problems is reduced considerably if
The dominating process during stage I is the formation and initiation of monomer-filled soap micelles. This stage corresponds with 10-20% of the monomer addition and determines the number of particles formed. This stage is also important for the properties of the latex since it (indirectly) determines the size of particles present in the final product. The main process in stage II is the growth of particles and in stage III is a diminishing monomer concentration the most important variable. Stabilization of the hydrophobic particles is crucial in all three stages. This can be achieved by either electrostatic repulsion of ionic groups at the polymer/water interface or steric stabilization by hydrophilic polymers (protective colloids), which can generate a so-called protective water barrier around the latex particle (Figure 2). Electrostatic stabilization is complementary to steric stabilization and the two are therefore frequently combined to achieve an optimum result [3,4].
Figure 2:A latex particle covered with degraded starch fragments.
The global demand for emulsion polymers (i.e. latexes) is increasing and studies show that this will continue in the near future (Table 3). The Asia/Pacific region has become the leading regional latex market in the last decade due to the very fast growing economies in India and China. This considerable growth is believed to continue in the years to come whilst the markets in North America and Western Europe are believed to increase at a subpar pace. Positive prospects are also expected in the regions Africa and Middle East.The predicted expansion of the latex market is partly based on the world wide desire to use more environmental benign products. The focus in developing nations is on the replacement of organic solvent-based products by their water-based counterparts whilst the industrialized world, due to the fact that waterborne products are already very common, is focusing on preparing them in a more sustainable way.
Protective water barrier
Particle
Protective water barrier
Introduction
14 15
Table 4:The Twelve Principles of Green Chemistry [9].
Principle Description
* Prevention It is better to prevent waste than to treat or clean up waste after it has been created.
* Atom Economy Synthetic methods should be designed to maximize the incorporation of all materials used in the process into the final product.
* Less Hazardous Syntheses Wherever practicable, synthetic methods should be designed to use and generate substances that possess little or no toxicity to human health and the environment.
* Designing Safer Chemicals Chemical products should be designed to affect their desired function while minimizing their toxicity.
* Safer Solvents and Auxiliaries The use of auxiliary substances (e.g., solvents, separation agents, etc.) should be made unnecessary wherever possible and innocuous when used.
* Energy Efficiency Energy requirements of chemical processes should be recognized for their environmental and economic impacts and should be minimized. If possible, synthetic methods should be conducted at ambient temperature and pressure.
* Renewable Feedstocks A raw material or feedstock should be renewable rather than depleting whenever technically and economically practicable.
* Reduce Derivatives Unnecessary derivatization (use of blocking groups, protection/ deprotection, temporary modification of physical/chemical processes) should be minimized or avoided if possible, because such steps require additional reagents and can generate waste.
* Catalysis Catalytic reagents (as selective as possible) are superior to stoichiometric reagents.
* Design for Degradation Chemical products should be designed so that at the end of their function they break down into innocuous degradation products and do not persist in the environment.
* Pollution Prevention Analytical methodologies need to be further developed to allow for real-time, in-process monitoring and control prior to the formation of hazardous substances.
* Accident Prevention Substances and the form of a substance used in a chemical process should be chosen to minimize the potential for chemical accidents, including releases, explosions, and fires.
this part of the formulation is made from similar raw materials as the additives used during the preparation of the latex. A total elimination of post-additions might be feasible with the use of a mixture of different protective colloids or a protective colloid that can be changed during the polymerization process. Different levels of emulsifier-like characteristics can be introduced to a (renewable) hydrophilic polymer by a hydrophobic derivatization before or during the preparation procedure [11]. The properties of the hydrophilic polymer can in some cases change considerably during the actual polymerization and reaction conditions can be used to fine tune the protective colloid characteristics (e.g. more or less degradation and grafting of monomer) during this part of the process. The example described above shows that the degree of sustainability of a synthetic latex can be improved according the twelve principles of green chemistry. However, an adaptation in the formulation can also have a considerable impact on the reaction conditions. Therefore, a change in the polymerization system (i.e. procedure and/or equipment) might be needed before the modified latex can be produced in an acceptable way. It is recommended to use the twelve principles of green engineering if changes in the polymerization system are actually needed (Table 5). Application of this guideline reduces the risk that the synthetic latex does not achieve its maximum level of sustainability possible.
Introduction
14 15
Table 4:The Twelve Principles of Green Chemistry [9].
Principle Description
* Prevention It is better to prevent waste than to treat or clean up waste after it has been created.
* Atom Economy Synthetic methods should be designed to maximize the incorporation of all materials used in the process into the final product.
* Less Hazardous Syntheses Wherever practicable, synthetic methods should be designed to use and generate substances that possess little or no toxicity to human health and the environment.
* Designing Safer Chemicals Chemical products should be designed to affect their desired function while minimizing their toxicity.
* Safer Solvents and Auxiliaries The use of auxiliary substances (e.g., solvents, separation agents, etc.) should be made unnecessary wherever possible and innocuous when used.
* Energy Efficiency Energy requirements of chemical processes should be recognized for their environmental and economic impacts and should be minimized. If possible, synthetic methods should be conducted at ambient temperature and pressure.
* Renewable Feedstocks A raw material or feedstock should be renewable rather than depleting whenever technically and economically practicable.
* Reduce Derivatives Unnecessary derivatization (use of blocking groups, protection/ deprotection, temporary modification of physical/chemical processes) should be minimized or avoided if possible, because such steps require additional reagents and can generate waste.
* Catalysis Catalytic reagents (as selective as possible) are superior to stoichiometric reagents.
* Design for Degradation Chemical products should be designed so that at the end of their function they break down into innocuous degradation products and do not persist in the environment.
* Pollution Prevention Analytical methodologies need to be further developed to allow for real-time, in-process monitoring and control prior to the formation of hazardous substances.
* Accident Prevention Substances and the form of a substance used in a chemical process should be chosen to minimize the potential for chemical accidents, including releases, explosions, and fires.
this part of the formulation is made from similar raw materials as the additives used during the preparation of the latex. A total elimination of post-additions might be feasible with the use of a mixture of different protective colloids or a protective colloid that can be changed during the polymerization process. Different levels of emulsifier-like characteristics can be introduced to a (renewable) hydrophilic polymer by a hydrophobic derivatization before or during the preparation procedure [11]. The properties of the hydrophilic polymer can in some cases change considerably during the actual polymerization and reaction conditions can be used to fine tune the protective colloid characteristics (e.g. more or less degradation and grafting of monomer) during this part of the process. The example described above shows that the degree of sustainability of a synthetic latex can be improved according the twelve principles of green chemistry. However, an adaptation in the formulation can also have a considerable impact on the reaction conditions. Therefore, a change in the polymerization system (i.e. procedure and/or equipment) might be needed before the modified latex can be produced in an acceptable way. It is recommended to use the twelve principles of green engineering if changes in the polymerization system are actually needed (Table 5). Application of this guideline reduces the risk that the synthetic latex does not achieve its maximum level of sustainability possible.
Introduction
16 17
Figure 3: The structure of amylose and AP.
Amylose is a mainly linear biopolymer and AP has a highly branched character. Linear segments originate from α-D-(1,4) linked AGU’s and branches are introduced by α-D-(1,6) linkages. Amylose and AP can differ considerably in chain length and number of α-D-(1,6) branches depending on the botanical origin. α-D-(1,4) AGU’s arrange into double helices (DH) and form tube like structures with a slightly hydrophobic interior which excludes the presence of water molecules. The inner size of the DH-tube is large enough to capture iodine molecules [13]. The DH-tubes are packed in the starch granule in a hexagonal way and these DH-tubes form crystalline lamellae (domains) in AP if they have enough freedom to align. DH-tubes remain amorphous in the vicinity of α-D-(1,6) linkages and in amylose. Two crystalline packing varieties are known which mainly differ in water content. The low water content packing (about 4 water molecules on 12 AGU’s) is designated as type A and can be found in plants evolved in a dry environment (e.g. cereal grain). Plants evolved in a more humid environment (e.g. potato) display a type B packing of the DH-tubes which contains approximately 36 molecules of water on 12 AGU’s. Figure 4 shows schematic representations of both hexagonal packings [14,15].
Figure 4: Schematic representation of type A and B packing of DH-tubes.
Amylose
Amylopectin
D-Glucose
Type - A Type - B
Double helices (DH) tube
Water
Table 5:The Twelve Principles of Green Engineering [10].Principle Description* Inherent Rather Than Circumstantial Designers need to strive to ensure that all materials
and energy inputs and outputs are as inherently nonhazardous as possible.
* Prevention Instead of Treatment It is better to prevent waste than to treat or clean up waste after it is formed.
* Design for Separation Separation and purification operations should be designed to minimize energy consumption and materials use.
* Maximize Efficiency Products, processes, and systems should be designed to maximize mass, energy, space, and time efficiency.
* Output-Pulled Versus Input-Pushed Products, processes, and systems should be "output pulled" rather than "input pushed" through the use of energy and materials
* Conserve Complexity Embedded entropy and complexity must be viewed as an investment when making design choices on recycle, reuse, or beneficial disposition.
* Durability Rather Than Immortality Targeted durability, not immortality, should be a design goal.
* Meet Need, Minimize Excess Design for unnecessary capacity or capability (e.g., "one size fits all") solutions should be considered a design flaw.
* Minimize Material Diversity Material diversity in multicomponent products should be minimized to promote disassembly and value retention.
* Integrate Material and Energy Flows Design of products, processes, and systems must include integration and interconnectivity with available energy and materials flows.
* Design for Commercial "Afterlife" Products, processes, and systems should be designed for performance in a commercial "afterlife."
* Renewable Rather Than Depleting Material and energy inputs should be renewable rather than depleting
StarchStarch is a carbohydrate based polymer only exceeded by cellulose in natural abundance. Starch is easier to disperse in water, and more reactive than cellulose and this renders starch more suitable for different kinds and degrees of modification [12]. Starch is therefore not only a good raw material because of its renewable nature, but also for its modification possibilities. It offers a high level of flexibility in order to design derivatives that fit a given application properly. The main building unit of starch is the α-D-glucopyranoside (D-glucose) molecule and the unit is called α-D-glucopyranosyl (α-glucan) or anhydroglucose unit (AGU) once the polysaccharide is formed by glycosidic bonds. The polysaccharide generated can be present as either amylose or amylopectin (AP) polymers (Figure 3).
Introduction
16 17
Figure 3: The structure of amylose and AP.
Amylose is a mainly linear biopolymer and AP has a highly branched character. Linear segments originate from α-D-(1,4) linked AGU’s and branches are introduced by α-D-(1,6) linkages. Amylose and AP can differ considerably in chain length and number of α-D-(1,6) branches depending on the botanical origin. α-D-(1,4) AGU’s arrange into double helices (DH) and form tube like structures with a slightly hydrophobic interior which excludes the presence of water molecules. The inner size of the DH-tube is large enough to capture iodine molecules [13]. The DH-tubes are packed in the starch granule in a hexagonal way and these DH-tubes form crystalline lamellae (domains) in AP if they have enough freedom to align. DH-tubes remain amorphous in the vicinity of α-D-(1,6) linkages and in amylose. Two crystalline packing varieties are known which mainly differ in water content. The low water content packing (about 4 water molecules on 12 AGU’s) is designated as type A and can be found in plants evolved in a dry environment (e.g. cereal grain). Plants evolved in a more humid environment (e.g. potato) display a type B packing of the DH-tubes which contains approximately 36 molecules of water on 12 AGU’s. Figure 4 shows schematic representations of both hexagonal packings [14,15].
Figure 4: Schematic representation of type A and B packing of DH-tubes.
Amylose
Amylopectin
D-Glucose
Type - A Type - B
Double helices (DH) tube
Water
Table 5:The Twelve Principles of Green Engineering [10].Principle Description* Inherent Rather Than Circumstantial Designers need to strive to ensure that all materials
and energy inputs and outputs are as inherently nonhazardous as possible.
* Prevention Instead of Treatment It is better to prevent waste than to treat or clean up waste after it is formed.
* Design for Separation Separation and purification operations should be designed to minimize energy consumption and materials use.
* Maximize Efficiency Products, processes, and systems should be designed to maximize mass, energy, space, and time efficiency.
* Output-Pulled Versus Input-Pushed Products, processes, and systems should be "output pulled" rather than "input pushed" through the use of energy and materials
* Conserve Complexity Embedded entropy and complexity must be viewed as an investment when making design choices on recycle, reuse, or beneficial disposition.
* Durability Rather Than Immortality Targeted durability, not immortality, should be a design goal.
* Meet Need, Minimize Excess Design for unnecessary capacity or capability (e.g., "one size fits all") solutions should be considered a design flaw.
* Minimize Material Diversity Material diversity in multicomponent products should be minimized to promote disassembly and value retention.
* Integrate Material and Energy Flows Design of products, processes, and systems must include integration and interconnectivity with available energy and materials flows.
* Design for Commercial "Afterlife" Products, processes, and systems should be designed for performance in a commercial "afterlife."
* Renewable Rather Than Depleting Material and energy inputs should be renewable rather than depleting
StarchStarch is a carbohydrate based polymer only exceeded by cellulose in natural abundance. Starch is easier to disperse in water, and more reactive than cellulose and this renders starch more suitable for different kinds and degrees of modification [12]. Starch is therefore not only a good raw material because of its renewable nature, but also for its modification possibilities. It offers a high level of flexibility in order to design derivatives that fit a given application properly. The main building unit of starch is the α-D-glucopyranoside (D-glucose) molecule and the unit is called α-D-glucopyranosyl (α-glucan) or anhydroglucose unit (AGU) once the polysaccharide is formed by glycosidic bonds. The polysaccharide generated can be present as either amylose or amylopectin (AP) polymers (Figure 3).
Introduction
18 19
Figure 5: Rapid Visco Analyzer (160 RPM) viscosity profiles of commercial available native starches (5 wt % in water)
Figure 6: Starch after modification by physical, (bio) chemical treatments and combinations thereof.
Physical and (bio) chemical modification are applied if the size and structure of the amylose and AP molecules needs to be changed for optimal performance in the application. However, the hydrophilic character of the amylose and AP molecules still dominates after these treatments and this can be changed by applying a derivatization of the starch before, during or after the modification step. The degree of substitution (DS) is defined as the number of substituted hydroxylic groups per AGU. An AGU contains three hydroxylic groups that are available for
Degradation & Recombination
Degradation & Debranching
Degradation & Elongation
Degradation & Elongationn
Amylopectin
Amylose
50
60
70
80
90
100
0
1000
2000
3000
4000
5000
0 2 4 6 8 10 12
Temp
eratu
re (°
C)
RVA
Visc
osity
(mPa
s)
Time (Minutes)
Potato
Waxy potato
Tapioca
Maize
Wheat
Temperature
The DH-tube packing has an effect on the dissolution characteristics. Type A is more difficult to dissolve than type B. Starch usually contains either type A or B packing of the DH-tubes (with the exception of pea starches that contain both). Type B packing of DH-tubes in the starch granule can be transformed into type A but this requires considerable dehydrative conditions. This conversion is irreversible even if the starch is stored with an amount of water above its original content [14-16]. The branches in AP can be classified as: A-chains, which are unsubstituted, B-chains, substituted by other chains and a C-chain corresponding to the original chain carrying the reducing glucose molecule [14]. A substantial variation in the position and length of the branches is observed between the starches of different botanical origin, which can result in considerable differences in properties, such as rheological behaviour. This behaviour is also strongly influenced by the amylose content of the starches and varieties without amylose are frequently preferred in industrial applications. AP (commonly referred to as waxy) starches are usually less sensitive to retrogradation and complexation to other substances in the formulation and require therefore a lower degree of derivatization to achieve the desired functionality. Among many root (e.g. tapioca), tuber (e.g potato) and cereal (e.g. maize) starches, the potato variety contains the lowest level of proteins and lipids. The structures of AP in root and tuber starches are very similar but differences can still be found. The AP part of potato starch contains, for example, more covalently bound phosphate groups than the tapioca counterpart and this helps to untangle potato AP during dissolution. This feature also reduces the occurence of partly dissolved starch granules for potato starch significantly and stabilizes the dissolved AP molecules [17].Native starch granules are insoluble in water because AP contains a considerable amount of crystalline domains. Enzymes involved in the synthesis of AP incorporate segments during synthesis which align to each other and form crystalline lamellae. These segments consist of double helices which have a tube like structure of more or less equal length [14]. A temperature of 60 °C or more, depending on the botanic origin, is needed to initiate swelling of the starch granules as the crystalline domains destabilize. A considerable increase in viscosity is observed during this process. The generated viscosity at this point is a combination of true viscosity and structure generated by swollen granules. The viscosity reduces in time due to disintegration of the swollen starch granules.A distinct increase in viscosity is observed during cooling. A continuous increase in viscosity, followed by gelling or precipitation is commonly seen for amylose containing native starches, a phenomenon called retrogradation. Each type of starch has a unique processing behaviour and this can be seen in the rapid visco analyzer (RVA) viscosity profiles of Figure 5. The viscosity behaviour of the native starches is not always desired and the starches are therefore frequently physically or (bio)chemically modified (Figure 6) [17-22].
Introduction
18 19
Figure 5: Rapid Visco Analyzer (160 RPM) viscosity profiles of commercial available native starches (5 wt % in water)
Figure 6: Starch after modification by physical, (bio) chemical treatments and combinations thereof.
Physical and (bio) chemical modification are applied if the size and structure of the amylose and AP molecules needs to be changed for optimal performance in the application. However, the hydrophilic character of the amylose and AP molecules still dominates after these treatments and this can be changed by applying a derivatization of the starch before, during or after the modification step. The degree of substitution (DS) is defined as the number of substituted hydroxylic groups per AGU. An AGU contains three hydroxylic groups that are available for
Degradation & Recombination
Degradation & Debranching
Degradation & Elongation
Degradation & Elongationn
Amylopectin
Amylose
50
60
70
80
90
100
0
1000
2000
3000
4000
5000
0 2 4 6 8 10 12
Temp
eratu
re (°
C)
RVA
Visc
osity
(mPa
s)
Time (Minutes)
Potato
Waxy potato
Tapioca
Maize
Wheat
Temperature
The DH-tube packing has an effect on the dissolution characteristics. Type A is more difficult to dissolve than type B. Starch usually contains either type A or B packing of the DH-tubes (with the exception of pea starches that contain both). Type B packing of DH-tubes in the starch granule can be transformed into type A but this requires considerable dehydrative conditions. This conversion is irreversible even if the starch is stored with an amount of water above its original content [14-16]. The branches in AP can be classified as: A-chains, which are unsubstituted, B-chains, substituted by other chains and a C-chain corresponding to the original chain carrying the reducing glucose molecule [14]. A substantial variation in the position and length of the branches is observed between the starches of different botanical origin, which can result in considerable differences in properties, such as rheological behaviour. This behaviour is also strongly influenced by the amylose content of the starches and varieties without amylose are frequently preferred in industrial applications. AP (commonly referred to as waxy) starches are usually less sensitive to retrogradation and complexation to other substances in the formulation and require therefore a lower degree of derivatization to achieve the desired functionality. Among many root (e.g. tapioca), tuber (e.g potato) and cereal (e.g. maize) starches, the potato variety contains the lowest level of proteins and lipids. The structures of AP in root and tuber starches are very similar but differences can still be found. The AP part of potato starch contains, for example, more covalently bound phosphate groups than the tapioca counterpart and this helps to untangle potato AP during dissolution. This feature also reduces the occurence of partly dissolved starch granules for potato starch significantly and stabilizes the dissolved AP molecules [17].Native starch granules are insoluble in water because AP contains a considerable amount of crystalline domains. Enzymes involved in the synthesis of AP incorporate segments during synthesis which align to each other and form crystalline lamellae. These segments consist of double helices which have a tube like structure of more or less equal length [14]. A temperature of 60 °C or more, depending on the botanic origin, is needed to initiate swelling of the starch granules as the crystalline domains destabilize. A considerable increase in viscosity is observed during this process. The generated viscosity at this point is a combination of true viscosity and structure generated by swollen granules. The viscosity reduces in time due to disintegration of the swollen starch granules.A distinct increase in viscosity is observed during cooling. A continuous increase in viscosity, followed by gelling or precipitation is commonly seen for amylose containing native starches, a phenomenon called retrogradation. Each type of starch has a unique processing behaviour and this can be seen in the rapid visco analyzer (RVA) viscosity profiles of Figure 5. The viscosity behaviour of the native starches is not always desired and the starches are therefore frequently physically or (bio)chemically modified (Figure 6) [17-22].
Introduction
20 21
water phase. This kind of particle formation is called homogeneous nucleation (Figure 8). The addition of surfactants and/or non-ionic amphiphiles is therefore not a necessity for a vinyl acetate emulsion polymerization. Reaction conditions can even be tuned to generate enough oligomeric radicals to replace the ionic surfactants and non-ionic amphiphiles completely [27-30].
Figure 8: A schematic representation of the homogeneous nucleation mechanism.
The high reactivity of the vinyl acetate radicals result in various transfer reactions to other components in the reaction mixture. It is even possible that the radical is transferred to a monomer which results in a mobile and rather stable monomeric free radical. This process is called desorption when the generated radical migrates out of the latex particle [5]. Radical termination occurs when a radical collides with another radical or a molecule with radical scavenging properties [31,32]. Chain transfer to a polymer is also possible and involves an H-abstraction of a polymer chain. Branching of the polyvinyl acetate polymer finds its origin in this type of radical formation and also grafting of hydrocarbon or hydroxyl based ingredients. This behaviour offers an interesting possibility for introducing water-soluble material as part of the latex particle by turning them into amphiphilics during the actual polymerization [33-35]. This process does firstly result in material that can replace the detergent and emulsifier part of the formulation and secondly in improved protective colloid characteristics. Vinyl acetate latexes that rely on this kind of stabilization have excellent freeze thaw properties and have viscosity levels and particle sizes which are out of reach with combinations of regular detergents and emulsifiers. The use of starch derivatives in vinyl acetate based polymerizations is known and frequently
M R O R M R O
MM O M
MMP MPMMP MPM
PMP PPPPM PPPMP MPM
Legend:
- M = Monomer; R = Radical; O = Oligomer; P = Polymer.
- Dashed circle: Amphiphilic part of polymer.
Stage I: Nucleation
Stage II: Growth Stage III: Consumption
Polym
eriza
tion r
ate
Monomer conversion (%) 0 100
derivatization and the DS can therefore not exceed three. An actual DS of three is not feasible due to differences in reactivity between the three hydroxylic groups and steric hindrance. Fortunately, there is no need to go this high because distinct changes in characteristics can usually already be observed around a DS of 0.01. The maximum DS obtainable during a given modification reaction (DSmax) is defined as the amount of reagent added with respect to the amount of AGU units present in the starch. The actual DS is lower than this value to an extent depending on the efficiency of the reaction in question. Etherification and esterification are commonly applied derivatization reactions (Figure 7).
Figure 7: The three reactive hydroxyl groups (R) of an AGU and three examples of possible derivatizations.
A hydroxyl group containing substituent (e.g. 2-hydroxylpropyl) competes with the hydroxyl groups of starch and makes the bonding of an increased amount of reagent to a AGU possible by the addition of reagent molecules to the subsituents already attached to the AGU. This type of derivatization can also be described in molar substitution (MS). This parameter is calculated in the same way as DS but is not limited to the value of 3.0 [14].
Modified starch in polyvinyl acetate based latexesVinyl acetate based latexes are extensively used in adhesives and coatings and the corresponding commercial markets have considerable growth prospects for the coming years [1,2,6]. A vinyl acetate polymerization reaction deviates significantly from the general rule of the Smith-Ewart theory that nucleation takes place mainly in monomer swollen micelles (in contrast to its styrene counterpart that is often used as reference system) [23-26]. This fundamental difference limits the applicability of the abundantly available information on “Smith-Ewart” based polymerizations. The origin is a higher water solubility of the vinyl acetate monomer compared to styrene.The hydrophilic nature of vinyl acetate allows formation of oligomeric radicals with up to ~10 units of monomer before precipitation occurs from the
OO
O
R
R
R
2
3
4
5
6
1
anhydroglucose unit (AGU)
R= Ether R= Ester R= Ester
1-Octenyl-succinate
(H)OOC
O
OStarch
StarchO
OH
2-Hydroxypropyl
StarchOO O
O(H)Starch
Phosphate
P
Introduction
20 21
water phase. This kind of particle formation is called homogeneous nucleation (Figure 8). The addition of surfactants and/or non-ionic amphiphiles is therefore not a necessity for a vinyl acetate emulsion polymerization. Reaction conditions can even be tuned to generate enough oligomeric radicals to replace the ionic surfactants and non-ionic amphiphiles completely [27-30].
Figure 8: A schematic representation of the homogeneous nucleation mechanism.
The high reactivity of the vinyl acetate radicals result in various transfer reactions to other components in the reaction mixture. It is even possible that the radical is transferred to a monomer which results in a mobile and rather stable monomeric free radical. This process is called desorption when the generated radical migrates out of the latex particle [5]. Radical termination occurs when a radical collides with another radical or a molecule with radical scavenging properties [31,32]. Chain transfer to a polymer is also possible and involves an H-abstraction of a polymer chain. Branching of the polyvinyl acetate polymer finds its origin in this type of radical formation and also grafting of hydrocarbon or hydroxyl based ingredients. This behaviour offers an interesting possibility for introducing water-soluble material as part of the latex particle by turning them into amphiphilics during the actual polymerization [33-35]. This process does firstly result in material that can replace the detergent and emulsifier part of the formulation and secondly in improved protective colloid characteristics. Vinyl acetate latexes that rely on this kind of stabilization have excellent freeze thaw properties and have viscosity levels and particle sizes which are out of reach with combinations of regular detergents and emulsifiers. The use of starch derivatives in vinyl acetate based polymerizations is known and frequently
M R O R M R O
MM O M
MMP MPMMP MPM
PMP PPPPM PPPMP MPM
Legend:
- M = Monomer; R = Radical; O = Oligomer; P = Polymer.
- Dashed circle: Amphiphilic part of polymer.
Stage I: Nucleation
Stage II: Growth Stage III: Consumption
Polym
eriza
tion r
ate
Monomer conversion (%) 0 100
derivatization and the DS can therefore not exceed three. An actual DS of three is not feasible due to differences in reactivity between the three hydroxylic groups and steric hindrance. Fortunately, there is no need to go this high because distinct changes in characteristics can usually already be observed around a DS of 0.01. The maximum DS obtainable during a given modification reaction (DSmax) is defined as the amount of reagent added with respect to the amount of AGU units present in the starch. The actual DS is lower than this value to an extent depending on the efficiency of the reaction in question. Etherification and esterification are commonly applied derivatization reactions (Figure 7).
Figure 7: The three reactive hydroxyl groups (R) of an AGU and three examples of possible derivatizations.
A hydroxyl group containing substituent (e.g. 2-hydroxylpropyl) competes with the hydroxyl groups of starch and makes the bonding of an increased amount of reagent to a AGU possible by the addition of reagent molecules to the subsituents already attached to the AGU. This type of derivatization can also be described in molar substitution (MS). This parameter is calculated in the same way as DS but is not limited to the value of 3.0 [14].
Modified starch in polyvinyl acetate based latexesVinyl acetate based latexes are extensively used in adhesives and coatings and the corresponding commercial markets have considerable growth prospects for the coming years [1,2,6]. A vinyl acetate polymerization reaction deviates significantly from the general rule of the Smith-Ewart theory that nucleation takes place mainly in monomer swollen micelles (in contrast to its styrene counterpart that is often used as reference system) [23-26]. This fundamental difference limits the applicability of the abundantly available information on “Smith-Ewart” based polymerizations. The origin is a higher water solubility of the vinyl acetate monomer compared to styrene.The hydrophilic nature of vinyl acetate allows formation of oligomeric radicals with up to ~10 units of monomer before precipitation occurs from the
OO
O
R
R
R
2
3
4
5
6
1
anhydroglucose unit (AGU)
R= Ether R= Ester R= Ester
1-Octenyl-succinate
(H)OOC
O
OStarch
StarchO
OH
2-Hydroxypropyl
StarchOO O
O(H)Starch
Phosphate
P
Introduction
22 23
process characteristics of a polymerization with an enzymatic converted starch (maltodextrin) as the only additive.Pyrodextrinated starch is already available for decades and the protective colloid of choice if high dry matter latexes are desired with high starch content. Chapter 3 shows the impact of changes in reaction conditions (i.e. hydrochloric acid concentration, level of pre-drying and dextrination time) on the properties of the pyrodextrins obtained and their protective colloid characteristics. Vinyl acetate based polymerizations can vary considerably in energy consumption during processing. This is because vinyl acetate can form a (low boiling or positive) azeotrope with water and the fact that polymerizations are frequently executed above the boiling point of this azeotrope. Monomer addition and radical formation therefore need to be carefully balanced to avoid excessive refluxing (c.q. heat loss). Chapter 4 deals with the effect of reaction temperature and initiator concentration on the product and process characteristics of a maltodextrin stabilized polyvinyl acetate latex. Modified starches with hydrophobic groups influence the particle formation process during the polymerization. The effect of different degrees of octenyl succinylation for both regular and waxy potato starch based derivatives is reported in chapter 5. The degree of octenyl succinylation of the starch is correlated with the viscosity of the latex and a number of the products prepared might be interesting for use as wood adhesives. The investigation described in chapter 6 is arranged around the standard wood adhesive test “EN204 D2” and this test is representative for wood glue application in areas with only limited exposure to moist (e.g kitchen or bathroom). The evaluated latexes will give an impression of the influence of an octenyl succinylation on wood bonding strength and the impact of using regular and waxy potato starch as raw material
AbbreviationsIUPAC : International union of pure and applied chemistry.AGU : Anhydroglucose unit.AP : Amylopectin.DH : Double helices.A-packing : Hexagonal packing of DH-tubes with about 4 water molecules on 12 AGU’sB-packing : Hexagonal packing of DH-tubes with about 36 water molecules on 12 AGU’sA-chains : AP-chain without substitution.B-chains : AP-chain substituted by other chains.C-chains : AP-chain carrying the reducing glucose molecule.DS : Degree of substitution.MS : Molar substitution.
patented [36-47]. Examples of the use of undegraded starch in the formulation can be found, but the application of smaller molecular size protective colloids is often favoured. The actual reason for this preference is not found in literature. The use of detergents and emulsifiers of petrochemical origin is almost allways claimed in these patents and products which require this type of additives rely on the combined effect of the starch and petrochemical based amphiphiles. Removing the petrochemical part of the formulation requires a considerable redesign in the preparation procedure to compensate for the change in reaction conditions. This change also offers a chance to evaluate the actual need of breaking the starch down to oligosaccharide level before the polymerization is started.
Scope of this thesisModified potato starch appears to be a good choice to replace synthetic additives (i.e. detergents, emulsifiers and protective colloids) in products prepared by free radical polymerizations. These additives are usually present in the range of 1 to 15 wt % (with respect to the amount of polymer) and a typical vinyl acetate based formulation contains approximately 10 wt % of these synthetic additives [48]. Patents about the use of modified starch as protective colloid in free radical polymerizations are not difficult to find and can be dated back at least for half a century [36-47]. However, the general attitude towards chemical products and processes changed considerably in the past decade. Nowadays, consumers want to use environmental benign products and this has never been an important issue in the past. This will be a positive drive for industries to increase the level of sustainability of the current products and preparation processes. The principles of green chemistry and engineering appear to be good guidelines to achieve this in an economical acceptable way.Raw materials that require only minimal modifications are preferred in green chemistry and the use of waxy (i.e. amylopectin) starch as raw material is therefore recommended. First of all, this type of starch is very low in amylose and requires less (chemical) modification for dissolution in water than its amylose containing counterparts. Root (tapioca) and tuber (potato) starches are favoured over other types of starches due to their lower content of proteins and lipids. Waxy potato starch contains more bound phosphate groups than its tapioca counterpart and in general this is not only beneficial with respect to dissolution characteristics, but to the stability of the solution as well. Potato starches are preferred as raw material in Europe over their tapioca based counterparts from a logistics point of view. Tapioca roots are grown in South America and Asia and need to be imported whilst potatoes are cultivated on a large scale in Europe. Furthermore, potatoes are cultivated and refined in a very organized and systematic way whilst tapioca roots are not. The variation in quality between batches of tapioca starch is therefore most likely much larger than that of potato starch. Finally, waxy potato starch is commercially available on a large scale whilst waxy tapioca is not.
This thesis investigates the potentials and limitations of (waxy) potato based protective colloids as alternatives for the commonly used synthetic additives in free radical polymerization. Chapter 2 describes the selected polymerization system and characterization methods to investigate this aspect for vinyl acetate based latexes. This chapter also shows the impact of changes in level of agitation and pre-dosages of monomer and initiator on product and
Introduction
22 23
process characteristics of a polymerization with an enzymatic converted starch (maltodextrin) as the only additive.Pyrodextrinated starch is already available for decades and the protective colloid of choice if high dry matter latexes are desired with high starch content. Chapter 3 shows the impact of changes in reaction conditions (i.e. hydrochloric acid concentration, level of pre-drying and dextrination time) on the properties of the pyrodextrins obtained and their protective colloid characteristics. Vinyl acetate based polymerizations can vary considerably in energy consumption during processing. This is because vinyl acetate can form a (low boiling or positive) azeotrope with water and the fact that polymerizations are frequently executed above the boiling point of this azeotrope. Monomer addition and radical formation therefore need to be carefully balanced to avoid excessive refluxing (c.q. heat loss). Chapter 4 deals with the effect of reaction temperature and initiator concentration on the product and process characteristics of a maltodextrin stabilized polyvinyl acetate latex. Modified starches with hydrophobic groups influence the particle formation process during the polymerization. The effect of different degrees of octenyl succinylation for both regular and waxy potato starch based derivatives is reported in chapter 5. The degree of octenyl succinylation of the starch is correlated with the viscosity of the latex and a number of the products prepared might be interesting for use as wood adhesives. The investigation described in chapter 6 is arranged around the standard wood adhesive test “EN204 D2” and this test is representative for wood glue application in areas with only limited exposure to moist (e.g kitchen or bathroom). The evaluated latexes will give an impression of the influence of an octenyl succinylation on wood bonding strength and the impact of using regular and waxy potato starch as raw material
AbbreviationsIUPAC : International union of pure and applied chemistry.AGU : Anhydroglucose unit.AP : Amylopectin.DH : Double helices.A-packing : Hexagonal packing of DH-tubes with about 4 water molecules on 12 AGU’sB-packing : Hexagonal packing of DH-tubes with about 36 water molecules on 12 AGU’sA-chains : AP-chain without substitution.B-chains : AP-chain substituted by other chains.C-chains : AP-chain carrying the reducing glucose molecule.DS : Degree of substitution.MS : Molar substitution.
patented [36-47]. Examples of the use of undegraded starch in the formulation can be found, but the application of smaller molecular size protective colloids is often favoured. The actual reason for this preference is not found in literature. The use of detergents and emulsifiers of petrochemical origin is almost allways claimed in these patents and products which require this type of additives rely on the combined effect of the starch and petrochemical based amphiphiles. Removing the petrochemical part of the formulation requires a considerable redesign in the preparation procedure to compensate for the change in reaction conditions. This change also offers a chance to evaluate the actual need of breaking the starch down to oligosaccharide level before the polymerization is started.
Scope of this thesisModified potato starch appears to be a good choice to replace synthetic additives (i.e. detergents, emulsifiers and protective colloids) in products prepared by free radical polymerizations. These additives are usually present in the range of 1 to 15 wt % (with respect to the amount of polymer) and a typical vinyl acetate based formulation contains approximately 10 wt % of these synthetic additives [48]. Patents about the use of modified starch as protective colloid in free radical polymerizations are not difficult to find and can be dated back at least for half a century [36-47]. However, the general attitude towards chemical products and processes changed considerably in the past decade. Nowadays, consumers want to use environmental benign products and this has never been an important issue in the past. This will be a positive drive for industries to increase the level of sustainability of the current products and preparation processes. The principles of green chemistry and engineering appear to be good guidelines to achieve this in an economical acceptable way.Raw materials that require only minimal modifications are preferred in green chemistry and the use of waxy (i.e. amylopectin) starch as raw material is therefore recommended. First of all, this type of starch is very low in amylose and requires less (chemical) modification for dissolution in water than its amylose containing counterparts. Root (tapioca) and tuber (potato) starches are favoured over other types of starches due to their lower content of proteins and lipids. Waxy potato starch contains more bound phosphate groups than its tapioca counterpart and in general this is not only beneficial with respect to dissolution characteristics, but to the stability of the solution as well. Potato starches are preferred as raw material in Europe over their tapioca based counterparts from a logistics point of view. Tapioca roots are grown in South America and Asia and need to be imported whilst potatoes are cultivated on a large scale in Europe. Furthermore, potatoes are cultivated and refined in a very organized and systematic way whilst tapioca roots are not. The variation in quality between batches of tapioca starch is therefore most likely much larger than that of potato starch. Finally, waxy potato starch is commercially available on a large scale whilst waxy tapioca is not.
This thesis investigates the potentials and limitations of (waxy) potato based protective colloids as alternatives for the commonly used synthetic additives in free radical polymerization. Chapter 2 describes the selected polymerization system and characterization methods to investigate this aspect for vinyl acetate based latexes. This chapter also shows the impact of changes in level of agitation and pre-dosages of monomer and initiator on product and
Introduction
24 25
59 (1988) 1151[29] W.D. Hergeth, W. Lebek, R. Kakuschke, K Schmutzler, Particle formation in emulsion polymerization, 1
Oligomers in emulsion polymerization of vinyl acetate, Makromol. Chem,, 192 (1991) 2265-2275.[30] C.D. Anderson, E.S. Daniels, Emulsion Polymerisation and Applications of Latex, Rapra Review Reports,
Volume 14, Number 4, 2003, Report 160[31] Sartomer Company, The effects polymerization inhibitors have on acrylate monomers and formulation
(http://www.docstoc.com/docs/78113538/The-Effects-Polymerization-Inhibitors-Have-On-Acrylate-Monomers; 20-06-2012)
[32] L.R. Bennedsen J.Muff, E.G. Sogaard influence of chloride and carbonates on the reactivity of activated persulfate, Chemosphere, 86 (2012) 1092-1097
[33] D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837.
[34] H. Lange, Emulsion polymerization of vinyl acetate with renewable raw materials as protective colloids, (http://kth.diva-portal.org/smash/record.jsf?pid=diva2:443050; 21-12-2012).
[35] M. Bödiger, S. Demharter, R. Mülhaupt, Starch and Dextrins in emulsion copolymerization, in: H. van Bekkum, H. Röper, F. Vorgagen, (Eds.), Carbohydrates as organic raw materials III , Wiley-VCH Verlag GmbH, Weinheim, 1996, pp141-154
[36] S.M. Hurley, F.L. Toss, P.E. Sandvick, S.E. Danley, Starch degradation/graft polymerization composition, process, and uses thereof, (2000) US6090884
[37] G.S. Samaranayake, R.T. Tomko, P.J. Ruhoff, M. Rao, Starch hybrid polymers (2011) WO2011008272A1[38] H. Büsching, K. Friederich, H. Buxhover, E. Abrahams, R. Gossen, W. Schaper, Adhesive dispersion for
gumming in envelope machines, WO/1998/011171[39] H. Buxhoffer, E. Abrahams-Meyer, R. Gossen, H. Büsching, K. Friederich, Rubber adhesive based on a
stablized polyvinyl acetate dispersion, WO2000027943.[40] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum
adhesive based on a filled polymer dispersion, WO2006094594A1.[41] V.A. Lauria, Remoistenable adhesive compositions (1987) US4678824[42] M.S. Mahiel, J.M. Cruden, Adhesive compositions and self-adhesive sheet materials (1988) EP0297900A2[43] M.S. Mahiel, J.M. Cruden, Surface coating compositions (1989) EP0351193A2[44] Jr.J. Wieczorek, L.M. Mahony, Aqueous adhesive compositions for use in binding book (1996) US5519072[45] U. Geissler, H. Hintz, U. Vogt-saggau, Powdery adhesive composition (1997) EP0799876A2[46] T. Mayer, H.P. Weitzel, R. Haerschel, T. Bastelberger, Method for producing polymers stabilized with
protective colloids (2000) US6300403[47] D.J. Guest, Manufacturing of polymeric material (1954) GB829149[48] T. Zanetta, F. Chiozza, C. Rei, Wood adhesive compositions (2013) WO2013057214A1
References[1] R.G. Gilbert, Emulsion Polymerization; A mechanistic approach, Academic press Limited, San Diego 1995[2] H. B. Yamak, Emulsion Polymerization: Effects of Polymerization Variables on the Properties of Vinyl
Acetate [3] C.S. Chern, Emulsion polymerization mechanisms and kinetics, Prog. Polym. Sci. 31 (2006) 443-486
[4] D. Urban, K. Takamura, Polymer dispersions and their industrial applications, Wiley-VHC Verlagh GmbH & Co, Weinheim, 2002.
[5] Freedoniagroup, World Emulsion Polymers: Industry Study with Forecasts for 2016 & 2021 (Executive summary), 2012
[6] Freedoniagroup, World Emulsion Polymers: Industry Study with Forecasts for 2016 & 2021, 2012 (http://www.freedoniagroup.com/brochure/29xx/2929smwe.pdf; 10-10-2012)
[7] J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical engineering: Green engineering, Chemical Engineering Journal, 133 ( 2007) 7-30
[8] M. Demendonca, T.E. Baxter, Design for the environment (DFE), An approach to achieve the ISO-14000 international standardization, Environmental management and health, 12 (1) (2001) 51-56
[9] ACS Green Chemistry Institute, The twelve principles of green chemistry,(www.acs.org; 18-03-2013)[10] ACS Green Chemistry Institute, The twelve principles of green engineering,( www.acs.org;18-03-2013)[11] H.J. De Vries, C. Semeijn, P.L. Buwalda, Emulsifier (2005) EP1743693[12] S. Richardson, L. Gorton, Characterisation of the substituent distribution in starch and cellulose derivatives,
Analytica Chimica Acta, 497 (2003) 27-65[13] K.A. Murdoch, The amylose-iodine complex, Carbohydrate Research, 233 (1992) 161-174[14] J. N. BeMiller, R.L. Whistler: Starch: Chemistry and Technology, third edition, Food science and technology,
International series, Academic Press, Burlington, 2009[15] A. Imberty, A. Buléon, V. Tran, S. Perez, Recent advances in knowledge of starch structure, 43 (10) (1991)
375-384[16] T.Y. Bogracheva, Y.L. Wang, T.L. Wang,C.L. Hedley, Structural studies of starches with different water
contents, Biopolymers, 64 (2002) 268-281[17] J.J.M. Swinkels, Composition and properties of commercial native starches, Starch, 37 (1) (1987) 1-5[18] G.O. Phillips, P.A. Villiams, Handbook of hydrocolloids, Woodhead Publishing Limited, Abington, 2005[19] H. Rubsam, M. Krottenhaler, M. Gastl,T. Becker, An overview of separation methods in starch analysis: The
importance of size exclusion chromatography and field flow fractionation Starch, 64 (9) (2012) 683-695[20] A.M. Hermansson, K. Svegmark, Developments in the understanding of starch functionality, Trends in food
science & technology, 7 (1996) 345[21] R. Hoover, Composition, molecular structure, and physicochemical properties of tuber and root starches: A
review, Carbohydrate polymers, 45 (2001) 253-267[22] R.M. Van den Einde, A.J. Van der Goot, R.M. Boom, Understanding molecular weight reduction of starch
during heating-shearing processes, Journal of food science, 68 (8) (2003) 2396-2404[23] D.C. Blackley, Emulsion Polymerization: Theory and practice, Applied Science Publishers LTD, London,
1975[24] C.S. Chern, G.W. Poehlein, Reaction kinetics of vinyl acetate emulsion polymerization, J. Appl. Polym. Sci.
33 (1987) 2117-2136.[25] S. Okamura, T. Motoyama, Emulsion polymerization of vinyl acetate, J. Polym. Sci. 58 (1962) 221-227.[26] J. Ugelstad, F.K. Hansen, Kinetics and mecahanism of emulsion polymerization, Rubber Chem. Techn, 49
(1976) 536-609.[27] H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/
bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).[28] T. Kourti, J.F. MacGregor, A.E. Hamielec, A study of particle nucleation and growth in the absence of soap
micelles during the emulsion polymerization of vinyl acetate in a pilot plant reactor, Polym Mater Sci Eng,
Introduction
24 25
59 (1988) 1151[29] W.D. Hergeth, W. Lebek, R. Kakuschke, K Schmutzler, Particle formation in emulsion polymerization, 1
Oligomers in emulsion polymerization of vinyl acetate, Makromol. Chem,, 192 (1991) 2265-2275.[30] C.D. Anderson, E.S. Daniels, Emulsion Polymerisation and Applications of Latex, Rapra Review Reports,
Volume 14, Number 4, 2003, Report 160[31] Sartomer Company, The effects polymerization inhibitors have on acrylate monomers and formulation
(http://www.docstoc.com/docs/78113538/The-Effects-Polymerization-Inhibitors-Have-On-Acrylate-Monomers; 20-06-2012)
[32] L.R. Bennedsen J.Muff, E.G. Sogaard influence of chloride and carbonates on the reactivity of activated persulfate, Chemosphere, 86 (2012) 1092-1097
[33] D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837.
[34] H. Lange, Emulsion polymerization of vinyl acetate with renewable raw materials as protective colloids, (http://kth.diva-portal.org/smash/record.jsf?pid=diva2:443050; 21-12-2012).
[35] M. Bödiger, S. Demharter, R. Mülhaupt, Starch and Dextrins in emulsion copolymerization, in: H. van Bekkum, H. Röper, F. Vorgagen, (Eds.), Carbohydrates as organic raw materials III , Wiley-VCH Verlag GmbH, Weinheim, 1996, pp141-154
[36] S.M. Hurley, F.L. Toss, P.E. Sandvick, S.E. Danley, Starch degradation/graft polymerization composition, process, and uses thereof, (2000) US6090884
[37] G.S. Samaranayake, R.T. Tomko, P.J. Ruhoff, M. Rao, Starch hybrid polymers (2011) WO2011008272A1[38] H. Büsching, K. Friederich, H. Buxhover, E. Abrahams, R. Gossen, W. Schaper, Adhesive dispersion for
gumming in envelope machines, WO/1998/011171[39] H. Buxhoffer, E. Abrahams-Meyer, R. Gossen, H. Büsching, K. Friederich, Rubber adhesive based on a
stablized polyvinyl acetate dispersion, WO2000027943.[40] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum
adhesive based on a filled polymer dispersion, WO2006094594A1.[41] V.A. Lauria, Remoistenable adhesive compositions (1987) US4678824[42] M.S. Mahiel, J.M. Cruden, Adhesive compositions and self-adhesive sheet materials (1988) EP0297900A2[43] M.S. Mahiel, J.M. Cruden, Surface coating compositions (1989) EP0351193A2[44] Jr.J. Wieczorek, L.M. Mahony, Aqueous adhesive compositions for use in binding book (1996) US5519072[45] U. Geissler, H. Hintz, U. Vogt-saggau, Powdery adhesive composition (1997) EP0799876A2[46] T. Mayer, H.P. Weitzel, R. Haerschel, T. Bastelberger, Method for producing polymers stabilized with
protective colloids (2000) US6300403[47] D.J. Guest, Manufacturing of polymeric material (1954) GB829149[48] T. Zanetta, F. Chiozza, C. Rei, Wood adhesive compositions (2013) WO2013057214A1
References[1] R.G. Gilbert, Emulsion Polymerization; A mechanistic approach, Academic press Limited, San Diego 1995[2] H. B. Yamak, Emulsion Polymerization: Effects of Polymerization Variables on the Properties of Vinyl
Acetate [3] C.S. Chern, Emulsion polymerization mechanisms and kinetics, Prog. Polym. Sci. 31 (2006) 443-486
[4] D. Urban, K. Takamura, Polymer dispersions and their industrial applications, Wiley-VHC Verlagh GmbH & Co, Weinheim, 2002.
[5] Freedoniagroup, World Emulsion Polymers: Industry Study with Forecasts for 2016 & 2021 (Executive summary), 2012
[6] Freedoniagroup, World Emulsion Polymers: Industry Study with Forecasts for 2016 & 2021, 2012 (http://www.freedoniagroup.com/brochure/29xx/2929smwe.pdf; 10-10-2012)
[7] J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical engineering: Green engineering, Chemical Engineering Journal, 133 ( 2007) 7-30
[8] M. Demendonca, T.E. Baxter, Design for the environment (DFE), An approach to achieve the ISO-14000 international standardization, Environmental management and health, 12 (1) (2001) 51-56
[9] ACS Green Chemistry Institute, The twelve principles of green chemistry,(www.acs.org; 18-03-2013)[10] ACS Green Chemistry Institute, The twelve principles of green engineering,( www.acs.org;18-03-2013)[11] H.J. De Vries, C. Semeijn, P.L. Buwalda, Emulsifier (2005) EP1743693[12] S. Richardson, L. Gorton, Characterisation of the substituent distribution in starch and cellulose derivatives,
Analytica Chimica Acta, 497 (2003) 27-65[13] K.A. Murdoch, The amylose-iodine complex, Carbohydrate Research, 233 (1992) 161-174[14] J. N. BeMiller, R.L. Whistler: Starch: Chemistry and Technology, third edition, Food science and technology,
International series, Academic Press, Burlington, 2009[15] A. Imberty, A. Buléon, V. Tran, S. Perez, Recent advances in knowledge of starch structure, 43 (10) (1991)
375-384[16] T.Y. Bogracheva, Y.L. Wang, T.L. Wang,C.L. Hedley, Structural studies of starches with different water
contents, Biopolymers, 64 (2002) 268-281[17] J.J.M. Swinkels, Composition and properties of commercial native starches, Starch, 37 (1) (1987) 1-5[18] G.O. Phillips, P.A. Villiams, Handbook of hydrocolloids, Woodhead Publishing Limited, Abington, 2005[19] H. Rubsam, M. Krottenhaler, M. Gastl,T. Becker, An overview of separation methods in starch analysis: The
importance of size exclusion chromatography and field flow fractionation Starch, 64 (9) (2012) 683-695[20] A.M. Hermansson, K. Svegmark, Developments in the understanding of starch functionality, Trends in food
science & technology, 7 (1996) 345[21] R. Hoover, Composition, molecular structure, and physicochemical properties of tuber and root starches: A
review, Carbohydrate polymers, 45 (2001) 253-267[22] R.M. Van den Einde, A.J. Van der Goot, R.M. Boom, Understanding molecular weight reduction of starch
during heating-shearing processes, Journal of food science, 68 (8) (2003) 2396-2404[23] D.C. Blackley, Emulsion Polymerization: Theory and practice, Applied Science Publishers LTD, London,
1975[24] C.S. Chern, G.W. Poehlein, Reaction kinetics of vinyl acetate emulsion polymerization, J. Appl. Polym. Sci.
33 (1987) 2117-2136.[25] S. Okamura, T. Motoyama, Emulsion polymerization of vinyl acetate, J. Polym. Sci. 58 (1962) 221-227.[26] J. Ugelstad, F.K. Hansen, Kinetics and mecahanism of emulsion polymerization, Rubber Chem. Techn, 49
(1976) 536-609.[27] H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/
bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).[28] T. Kourti, J.F. MacGregor, A.E. Hamielec, A study of particle nucleation and growth in the absence of soap
micelles during the emulsion polymerization of vinyl acetate in a pilot plant reactor, Polym Mater Sci Eng,
26
CHAPTER 2Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
26
CHAPTER 2Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
28 29
radicals formed [9,11,12]. As a consequence, buffering is frequently applied (e.g. with sodium bicarbonate) in order to maintain the pH (of the reaction mixture) typically below 7 in order to avoid excessive saponification of both VAM and the polymer.Emulsion polymerizations can be executed in batch, semi-batch and continuous mode. Continuous polymerization reactors (e.g oscillatory-baffled reactors) are emerging and are favoured with respect to size, flexibility and efficiency [13-15]. Conversely, a typical VAM polymerisation is still carried out in a semi-batch configuration most of the time and application of a continuous reactor is not a commonly accepted alternative up to now. Reaction in a semi-batch reactor can be tuned by changing dosage rates of reagents and additives. When taking into account heat transfer, reactors based on stainless steel are preferred above glass counterparts due to their higher thermal conductivity [16,17]. However, steel reactors are prone to metal leaching and this may affect the rate of radical initiation. Pickling, passivation or coatings are suitable treatments if an increased resistance to corrosive circumstances is desired [18,19]. The degree of mixing during processing, as common in all emulsification processes, is of paramount importance. A proper homogenisation can be achieved by laminar (low shear) or turbulent (high shear) flow-based mixing and combinations thereof. An efficient low shear mixing is desired during the polymerization and a helical ribbon configuration meets this criterion properly [20-28].The polymerization process (including experimental conditions, type of additives and reactor intakes) is intimately related to the final product properties. Viscosity (up to 105 mPa·s) and colloidal stability (over 6 months) are important properties of the latexes obtained and a direct function of the number and size of the particles obtained [29-35]. A monodisperse particle size distribution (PSD) is often desirable (and aimed at in this work) in connection with a proper understanding of structure-property relationships. The product properties (including latex viscosity and stability) are not only related to the type of protective colloid (pVOH) used (e.g. chain length and degree of saponification) but also to the number of pVAc chains grafted on the pVOH during the free radical polymerization [36,37]. Furthermore the VAM conversion is of crucial importance since the presence of unreacted monomer significantly affects latex viscosity and particle size [38,39]. From an application point of view, the minimum film formation temperature is an important property [40,41], which can be correlated with the glass transition temperature (Tg) of the latex in solution [42]. The application of a modulated DSC (mDSC) is often recommended for the Tg determination since interference from volume relaxation processes can be excluded by using the reversing heat flow based data [43-45].The demand for more sustainable chemical products is steadily increasing on a global scale and currently available products can be improved significantly by the use of renewable ingredients, optimization of reaction conditions, smart plant designs and combinations thereof [46,47]. The principles of green chemistry and engineering represent useful guidelines, including the recommended application of renewable raw materials [48-50]. As a consequence, there is a clear incentive to (partly) replace pVOH (in pVAc production) with green alternatives like hydroxyethyl cellulose, proteins and starch [37,51-53]. A typical pVAc formulation contains about 10 wt % synthetic additives that might be replaced by renewable counterparts. In this context, the use of starch has been already reported. However, specific details about reaction conditions (i.e. dosage protocols, pH profile, reaction temperature profile, level of agitation, inhibitors presence and reactor configuration) are usually only partly described
AbstractThe free radical polymerization of vinyl acetate in the presence of a waxy potato starch-based maltodextrin as protective colloid in the absence of detergents, emulsifiers and anti-foaming agents, was investigated systematically. A simplified Taguchi design of experiments was used to determine the influence of the agitation level, pre-dosage of monomer and initiator/buffer mixture (constituting three independent factors) and their interactions on process conditions and latex characteristics. Several responses (i.e. heat flow during processing, monomer conversion, product recovery, anion concentration, pH, viscosity, particle size distribution, amount of grafted protective colloid and glass transition temperature) were determined and related with the three independent factors described above. The reproducibility level of the system was good as evident from the statistical analysis. A number of significant effects and correlations were found, which constitute useful information to optimize the process and product characteristics of this type of latexes.
IntroductionThe widespread use of polyvinyl acetate-based polymers (pVAc) finds its origin in a combination of relatively low costs, superior performance and the fact that the properties of the polymer can be tuned in numerous ways to meet requirements for different applications (e.g. surface coatings, caulks and adhesives). PVAc exhibits excellent adhesion to cellulosic and other materials and are therefore used as adhesives in many different application areas. Besides excellent performance, processing and formulation technologies are also well developed [1]. The global demand for pVAc was around 1.2 million tons in 2005 and has grown ever since. This positive trend is believed to continue in the coming years due to the global trend to replace solvent borne products with water-based counterparts [2,3]. PVAc is commonly prepared by emulsion polymerization, which requires water, vinyl acetate monomer (VAM), a free radical initiator and, in most cases, additives including protective colloids, surfactants and emulsifiers. Protective colloids, consisting typically of soluble polymeric chains, are added in order to stabilize the polymer/monomer particles and thus to prevent coalescence. Among all possible choices, polyvinyl alcohol (pVOH) is commonly used as a protective colloid due to its availability and relatively low cost [4,5]. VAM polymerizations without emulsifiers are also possible (and industrially desirable) but only if sufficient oligomeric radicals are generated during the polymerization process [6]. These VAM based oligomeric radicals exhibit surfactant-like characteristics and, as such, they factually function as emulsifiers. Various free radical initiation systems have been developed for the synthesis of pVAc. A well known example is the use of persulfate, which decomposes into radicals by heating and/or in the presence of reducing agents [7-9]. Significant thermal dissociation requires temperatures above 50°C and a post-treatment with heat or reducing agent (e.g. sodium thiosulfate) in order to convert unreacted persulfate at the end of the polymerization. Potassium persulfate is frequently used in free radical polymerizations, despite the fact that it is less water soluble than the varieties based on sodium and ammonium [10]. The actual reason for this preference is not yet found in literature. The pH is a crucial factor in this polymerization process as it does not only affect the rate of radical formation, but also the ratio of sulfate and hydroxyl
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
28 29
radicals formed [9,11,12]. As a consequence, buffering is frequently applied (e.g. with sodium bicarbonate) in order to maintain the pH (of the reaction mixture) typically below 7 in order to avoid excessive saponification of both VAM and the polymer.Emulsion polymerizations can be executed in batch, semi-batch and continuous mode. Continuous polymerization reactors (e.g oscillatory-baffled reactors) are emerging and are favoured with respect to size, flexibility and efficiency [13-15]. Conversely, a typical VAM polymerisation is still carried out in a semi-batch configuration most of the time and application of a continuous reactor is not a commonly accepted alternative up to now. Reaction in a semi-batch reactor can be tuned by changing dosage rates of reagents and additives. When taking into account heat transfer, reactors based on stainless steel are preferred above glass counterparts due to their higher thermal conductivity [16,17]. However, steel reactors are prone to metal leaching and this may affect the rate of radical initiation. Pickling, passivation or coatings are suitable treatments if an increased resistance to corrosive circumstances is desired [18,19]. The degree of mixing during processing, as common in all emulsification processes, is of paramount importance. A proper homogenisation can be achieved by laminar (low shear) or turbulent (high shear) flow-based mixing and combinations thereof. An efficient low shear mixing is desired during the polymerization and a helical ribbon configuration meets this criterion properly [20-28].The polymerization process (including experimental conditions, type of additives and reactor intakes) is intimately related to the final product properties. Viscosity (up to 105 mPa·s) and colloidal stability (over 6 months) are important properties of the latexes obtained and a direct function of the number and size of the particles obtained [29-35]. A monodisperse particle size distribution (PSD) is often desirable (and aimed at in this work) in connection with a proper understanding of structure-property relationships. The product properties (including latex viscosity and stability) are not only related to the type of protective colloid (pVOH) used (e.g. chain length and degree of saponification) but also to the number of pVAc chains grafted on the pVOH during the free radical polymerization [36,37]. Furthermore the VAM conversion is of crucial importance since the presence of unreacted monomer significantly affects latex viscosity and particle size [38,39]. From an application point of view, the minimum film formation temperature is an important property [40,41], which can be correlated with the glass transition temperature (Tg) of the latex in solution [42]. The application of a modulated DSC (mDSC) is often recommended for the Tg determination since interference from volume relaxation processes can be excluded by using the reversing heat flow based data [43-45].The demand for more sustainable chemical products is steadily increasing on a global scale and currently available products can be improved significantly by the use of renewable ingredients, optimization of reaction conditions, smart plant designs and combinations thereof [46,47]. The principles of green chemistry and engineering represent useful guidelines, including the recommended application of renewable raw materials [48-50]. As a consequence, there is a clear incentive to (partly) replace pVOH (in pVAc production) with green alternatives like hydroxyethyl cellulose, proteins and starch [37,51-53]. A typical pVAc formulation contains about 10 wt % synthetic additives that might be replaced by renewable counterparts. In this context, the use of starch has been already reported. However, specific details about reaction conditions (i.e. dosage protocols, pH profile, reaction temperature profile, level of agitation, inhibitors presence and reactor configuration) are usually only partly described
AbstractThe free radical polymerization of vinyl acetate in the presence of a waxy potato starch-based maltodextrin as protective colloid in the absence of detergents, emulsifiers and anti-foaming agents, was investigated systematically. A simplified Taguchi design of experiments was used to determine the influence of the agitation level, pre-dosage of monomer and initiator/buffer mixture (constituting three independent factors) and their interactions on process conditions and latex characteristics. Several responses (i.e. heat flow during processing, monomer conversion, product recovery, anion concentration, pH, viscosity, particle size distribution, amount of grafted protective colloid and glass transition temperature) were determined and related with the three independent factors described above. The reproducibility level of the system was good as evident from the statistical analysis. A number of significant effects and correlations were found, which constitute useful information to optimize the process and product characteristics of this type of latexes.
IntroductionThe widespread use of polyvinyl acetate-based polymers (pVAc) finds its origin in a combination of relatively low costs, superior performance and the fact that the properties of the polymer can be tuned in numerous ways to meet requirements for different applications (e.g. surface coatings, caulks and adhesives). PVAc exhibits excellent adhesion to cellulosic and other materials and are therefore used as adhesives in many different application areas. Besides excellent performance, processing and formulation technologies are also well developed [1]. The global demand for pVAc was around 1.2 million tons in 2005 and has grown ever since. This positive trend is believed to continue in the coming years due to the global trend to replace solvent borne products with water-based counterparts [2,3]. PVAc is commonly prepared by emulsion polymerization, which requires water, vinyl acetate monomer (VAM), a free radical initiator and, in most cases, additives including protective colloids, surfactants and emulsifiers. Protective colloids, consisting typically of soluble polymeric chains, are added in order to stabilize the polymer/monomer particles and thus to prevent coalescence. Among all possible choices, polyvinyl alcohol (pVOH) is commonly used as a protective colloid due to its availability and relatively low cost [4,5]. VAM polymerizations without emulsifiers are also possible (and industrially desirable) but only if sufficient oligomeric radicals are generated during the polymerization process [6]. These VAM based oligomeric radicals exhibit surfactant-like characteristics and, as such, they factually function as emulsifiers. Various free radical initiation systems have been developed for the synthesis of pVAc. A well known example is the use of persulfate, which decomposes into radicals by heating and/or in the presence of reducing agents [7-9]. Significant thermal dissociation requires temperatures above 50°C and a post-treatment with heat or reducing agent (e.g. sodium thiosulfate) in order to convert unreacted persulfate at the end of the polymerization. Potassium persulfate is frequently used in free radical polymerizations, despite the fact that it is less water soluble than the varieties based on sodium and ammonium [10]. The actual reason for this preference is not yet found in literature. The pH is a crucial factor in this polymerization process as it does not only affect the rate of radical formation, but also the ratio of sulfate and hydroxyl
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
30 31
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM intake and dosage rate was also monitored with a balance. STS (as a 0.3 M solution), SPS (3 wt % with respect to the total weight) and SBC (4 wt %) were added according to the following. Two syringe pumps were used to add a premix of SPS and sodium bicarbonate (SBC). One was reserved for the pre-dosage and the other for the dosage during the polymerization run. A third syringe pump was used to add sodium thiosulfate (STS) after the actual polymerization was finished. Each syringe pump had the same dosage protocol in all experiments. Differences in dosages were achieved by changing the concentration of solutions used. This approach minimizes the variation in water content during the polymerization and this is crucial because the initial radical formation takes place in the water phase and is concentration dependent.
Figure 1. Schematic representation of the polymerization reactor.
The water bath used expresses the power consumption (WPC) in a percentage in the range of -100 to 100 % and this value was recorded every 3 seconds during the preparation procedure. The amount of water present in the water bath was approximately the same in each experiment in order to minimize the variation between the executed experiments. The sum of all recordings in the period from 0-7 hours was used as a measure for the consumed power during polymerization.
Statistical designA Taguchi L8 orthogonal design allows the evaluation of three factors (i.e. pre-dosage of monomer and initiator as well as agitation level) and their interactions on the selected responses [57,58]. The corresponding levels for every factor were designated as low (L) and high (H). The effect of a factor, or an interaction, was calculated by multiplying the measured value with the corresponding number in the Taguchi L8 matrix (Table 1). The effect of a factor, or an interaction, is defined as the sum of the obtained values of the eight runs performed after this multiplication [59]. The output is the effect of a factor, or interaction, on the selected response and is similar to the output of factorial design calculations. The Taguchi is augmented with a centre point (C) measured in triplicate for the estimation of the level of noise in the experiment.
or not mentioned at all [54-56]. Replacement of pVOH leads to pronounced differences in process and product characteristics. Consequently, detailed knowledge about the impact of the use of starch derivatives on process and product characteristics is essential for further process and product design.This investigation involves a pVAc synthesis in which a modified starch was used to replace not only the pVOH part of the formulation, but other synthetic additives (e.g. surfactants) as well. Waxy potato starch (i.e. amylopectin potato starch) is an excellent candidate as it is the most stable starch available after dissolution in water since the amylose content is close to negligible (no retrogradation) and the amylopectin part is partly stabilized by covalently bound phosphate. Moreover, the proteins and lipids content of waxy potato starch is very low, considerable lower than for waxy maize for example, which reduces the risk of foaming during processing significantly. However, potato starch dissolves with a considerable peak in viscosity and a degradation step (acidic, oxidative, mechanic or enzymatic) is required to avoid considerable differences in viscosity during polymerization. Enzymatically degraded waxy potato starch dissolves without a peak in viscosity and is therefore used at 10 wt % on pVAc as protective colloid in this study. The reaction mixture contained oxygen during start-up of the polymerization and the reaction was executed with hydroquinone containing VAM, thus resembling experimental settings more commonly applied at industrial level. The selected pVAc formulation had a final water/pVAc ratio of 1 (weight/weight), considered a minimum value to be commercially attractive. In particular, the influence of different processing variables (i.e. pre-dosage of monomer or initiator as well as agitation level) on process and product characteristics was investigated. This was achieved through the use of statistical analytical tool, i.e. a simplified Taguchi experimental design.
ExperimentalMaterialsThe modified waxy potato starch (moisture content: 6.4 wt %) used in this study is an enzymatically degraded starch commercially available under the trade name of Eliane MD2 (AVEBE U.A.). The vinyl acetate monomer (VAM, from ACROS) contains 3-30 ppm hydroquinone and was used as such. Analytical reagent grade sodium persulfate (SPS) supplied by VWR International. Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) both of analytical quality and retrieved from Merck. All ingredients were used without additional purification and in all cases the solvent was demineralised water.
EquipmentA double jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles with a width of 1 cm and an outer dimension of 10.5x7 cm (height x diameter)) was applied. A lid made of borosilicate glass with several connection points was placed on top of the reactor and the reactor was completely insulated with radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the temperature of the headspace in the reactor (HST). The feeding lines of VAM and the SPS/SBC mixture were placed outside the reflux region with the aid of a glass attachment, with multiple dosage points, to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 1).
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
30 31
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM intake and dosage rate was also monitored with a balance. STS (as a 0.3 M solution), SPS (3 wt % with respect to the total weight) and SBC (4 wt %) were added according to the following. Two syringe pumps were used to add a premix of SPS and sodium bicarbonate (SBC). One was reserved for the pre-dosage and the other for the dosage during the polymerization run. A third syringe pump was used to add sodium thiosulfate (STS) after the actual polymerization was finished. Each syringe pump had the same dosage protocol in all experiments. Differences in dosages were achieved by changing the concentration of solutions used. This approach minimizes the variation in water content during the polymerization and this is crucial because the initial radical formation takes place in the water phase and is concentration dependent.
Figure 1. Schematic representation of the polymerization reactor.
The water bath used expresses the power consumption (WPC) in a percentage in the range of -100 to 100 % and this value was recorded every 3 seconds during the preparation procedure. The amount of water present in the water bath was approximately the same in each experiment in order to minimize the variation between the executed experiments. The sum of all recordings in the period from 0-7 hours was used as a measure for the consumed power during polymerization.
Statistical designA Taguchi L8 orthogonal design allows the evaluation of three factors (i.e. pre-dosage of monomer and initiator as well as agitation level) and their interactions on the selected responses [57,58]. The corresponding levels for every factor were designated as low (L) and high (H). The effect of a factor, or an interaction, was calculated by multiplying the measured value with the corresponding number in the Taguchi L8 matrix (Table 1). The effect of a factor, or an interaction, is defined as the sum of the obtained values of the eight runs performed after this multiplication [59]. The output is the effect of a factor, or interaction, on the selected response and is similar to the output of factorial design calculations. The Taguchi is augmented with a centre point (C) measured in triplicate for the estimation of the level of noise in the experiment.
or not mentioned at all [54-56]. Replacement of pVOH leads to pronounced differences in process and product characteristics. Consequently, detailed knowledge about the impact of the use of starch derivatives on process and product characteristics is essential for further process and product design.This investigation involves a pVAc synthesis in which a modified starch was used to replace not only the pVOH part of the formulation, but other synthetic additives (e.g. surfactants) as well. Waxy potato starch (i.e. amylopectin potato starch) is an excellent candidate as it is the most stable starch available after dissolution in water since the amylose content is close to negligible (no retrogradation) and the amylopectin part is partly stabilized by covalently bound phosphate. Moreover, the proteins and lipids content of waxy potato starch is very low, considerable lower than for waxy maize for example, which reduces the risk of foaming during processing significantly. However, potato starch dissolves with a considerable peak in viscosity and a degradation step (acidic, oxidative, mechanic or enzymatic) is required to avoid considerable differences in viscosity during polymerization. Enzymatically degraded waxy potato starch dissolves without a peak in viscosity and is therefore used at 10 wt % on pVAc as protective colloid in this study. The reaction mixture contained oxygen during start-up of the polymerization and the reaction was executed with hydroquinone containing VAM, thus resembling experimental settings more commonly applied at industrial level. The selected pVAc formulation had a final water/pVAc ratio of 1 (weight/weight), considered a minimum value to be commercially attractive. In particular, the influence of different processing variables (i.e. pre-dosage of monomer or initiator as well as agitation level) on process and product characteristics was investigated. This was achieved through the use of statistical analytical tool, i.e. a simplified Taguchi experimental design.
ExperimentalMaterialsThe modified waxy potato starch (moisture content: 6.4 wt %) used in this study is an enzymatically degraded starch commercially available under the trade name of Eliane MD2 (AVEBE U.A.). The vinyl acetate monomer (VAM, from ACROS) contains 3-30 ppm hydroquinone and was used as such. Analytical reagent grade sodium persulfate (SPS) supplied by VWR International. Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) both of analytical quality and retrieved from Merck. All ingredients were used without additional purification and in all cases the solvent was demineralised water.
EquipmentA double jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles with a width of 1 cm and an outer dimension of 10.5x7 cm (height x diameter)) was applied. A lid made of borosilicate glass with several connection points was placed on top of the reactor and the reactor was completely insulated with radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the temperature of the headspace in the reactor (HST). The feeding lines of VAM and the SPS/SBC mixture were placed outside the reflux region with the aid of a glass attachment, with multiple dosage points, to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 1).
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
32 33
Initial experiments were performed at the CCC settings (RPM:VAM:SPS) to determine acceptable dosage rates for the VAM and SPS solutions during processing. Typical VAM polymerizations require 6-8 hours and the dosage time of VAM was therefore limited to approximately 5 hours. The SPS solution was dosed for a slightly longer period (6 hours) in order to allow the VAM stored in the latex particles to react as well. A polymerization procedure with only a modest refluxing level was obtained by carefully balancing the concentration of the SPS solution and the VAM dosage rate. The polymerization procedure was not directly stopped after the SPS addition was finished because the reaction mixture still contained significant amounts of SPS and the VAM stored in the latex particles was not yet consumed at this point (as evidenced by conversion measurements). The reaction time was elongated until the latex had a residual VAM content below 10 mg/g after the termination procedure (addition of STS and cooling down to 20°C).
CharacterizationViscosity, pH and dry matter of the latexes were determined with a Brookfield DV-II+(20 RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80 °C), respectively. Ethanal and residual VAM were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) gas and a flame ionization detector. The gas chromatograph measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 minutes and the supernatant was mixed (1:1 on volume) with 5 mM NaOH. This mixture was used to quantify the anion composition (sulfate, thiosulfate and (free) acetate) with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. PSD’s were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13-32.5 μm). Fraunhofer theory based calculations were used and the performed calculations are ISO 13320 compliant. The mode and the half width were the selected variables to describe the obtained PSD. Each PSD is an average of ten sequential measurements with an interval of 10 seconds. The amount of material smaller than 1 μm in the PSD can be used (in this particular case) as a measure of the sensitivity of particle coagulation during measuring and is defined as the difference between the maximum and minimum of 10 of these measurements performed (stability response).Determination of level of grafted material (between starch and pVAc) was based on a precipitation of 3-4 g dispersion in 100 ml acetone (0-5 °C) followed by soxhlet extraction in acetone. The soxhlet extraction was automated with a Soxtec 2043/2046 system from Foss Analytical (160°C; cellulose thimbles). The extraction involves 6 hours of boiling and subsequent 18 hours of rinsing. This is largely sufficient to remove the homopolymer, as checked with different times of rinsing of physical mixtures of starch and pVAc. The grafting efficiency (f) is defined as:
0
10
20
30
40
50
0
20
40
60
80
100
-1.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Mono
mer
(mmo
l/min)
Temp
eratu
re (°
C)
Time (hours) Profile waterbath Pre-dosage VAM: L Pre-dosage VAM: H Pre-dosage VAM: C
Water or SPS/SBC: 10ml (10ml/min) + SPS/SBC: 30ml (5ml/hr) STS 3ml (3ml/hr)
Table 1:Taguchi L8 matrix for three factors and their interactions. AT = RPM, BT = pre-dosage of VAM, CT = pre-dosage of SPS. The applied levels (L = low, H = high and C = centre) are specified below and in Table 2.
Run Code Factors or interactionsAT BT ATBT CT ATCT BTCT ATBTCT
1 LLL -1 -1 -1 -1 -1 -1 -12 LLH -1 -1 -1 1 1 1 13 LHL -1 1 1 -1 -1 1 14 LHH -1 1 1 1 1 -1 -15 HLL 1 -1 1 -1 1 -1 16 HLH 1 -1 1 1 -1 1 -17 HHL 1 1 -1 -1 1 1 -18 HHH 1 1 -1 1 -1 -1 1
Procedure The polymerization reactor was filled with 256.2 g of demineralized water and 35.6 g of enzymatic degraded waxy potato starch. Automatic mixing (levels = 90:120:150 RPM) was started after one minute of gentle manual homogenization of the starch/water mixture in the reactor with the stirrer. The applied temperature profile and dosage protocols of VAM, SPS and STS are given in Figure 2. A total of 0.3 kg VAM was added in all cases and the levels of pre-dosage are 0, 5 and 10 % of total added VAM, respectively. The actual dosage was monitored in time and the real added amount of VAM was used for mass balance calculations. 5.3 mmol SPS and 20.0 mmol SBC were dissolved in 30-40 ml demineralised water and added in all cases as a mixture. The actual addition of the mixture starts after 104 minutes with a pre-dosage of 10 ml (10 ml/min) followed by 30 ml with a dosage speed of 5 ml/hour. The water of the pre-dosage contains 0, 12.5 or 25 wt % of the total amount of SPS and SBC solution (and therefore the remaining 30 ml contains 100, 87.5 or 75 wt % respectively). 3 ml of a 0.3 M STS solution was added with 3 ml/hour after the water bath temperature drops significantly below 65°C during the cooling down. Agitation was continued for at least one hour after the temperature of the water bath reaches a temperature of 20 °C. The dispersion was transferred into a storage container without any additional treatments after this short period of equilibration.
Figure 2:Applied temperature profile and dosage protocols of VAM, SPS/SBC and STS.
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
32 33
Initial experiments were performed at the CCC settings (RPM:VAM:SPS) to determine acceptable dosage rates for the VAM and SPS solutions during processing. Typical VAM polymerizations require 6-8 hours and the dosage time of VAM was therefore limited to approximately 5 hours. The SPS solution was dosed for a slightly longer period (6 hours) in order to allow the VAM stored in the latex particles to react as well. A polymerization procedure with only a modest refluxing level was obtained by carefully balancing the concentration of the SPS solution and the VAM dosage rate. The polymerization procedure was not directly stopped after the SPS addition was finished because the reaction mixture still contained significant amounts of SPS and the VAM stored in the latex particles was not yet consumed at this point (as evidenced by conversion measurements). The reaction time was elongated until the latex had a residual VAM content below 10 mg/g after the termination procedure (addition of STS and cooling down to 20°C).
CharacterizationViscosity, pH and dry matter of the latexes were determined with a Brookfield DV-II+(20 RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80 °C), respectively. Ethanal and residual VAM were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) gas and a flame ionization detector. The gas chromatograph measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 minutes and the supernatant was mixed (1:1 on volume) with 5 mM NaOH. This mixture was used to quantify the anion composition (sulfate, thiosulfate and (free) acetate) with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. PSD’s were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13-32.5 μm). Fraunhofer theory based calculations were used and the performed calculations are ISO 13320 compliant. The mode and the half width were the selected variables to describe the obtained PSD. Each PSD is an average of ten sequential measurements with an interval of 10 seconds. The amount of material smaller than 1 μm in the PSD can be used (in this particular case) as a measure of the sensitivity of particle coagulation during measuring and is defined as the difference between the maximum and minimum of 10 of these measurements performed (stability response).Determination of level of grafted material (between starch and pVAc) was based on a precipitation of 3-4 g dispersion in 100 ml acetone (0-5 °C) followed by soxhlet extraction in acetone. The soxhlet extraction was automated with a Soxtec 2043/2046 system from Foss Analytical (160°C; cellulose thimbles). The extraction involves 6 hours of boiling and subsequent 18 hours of rinsing. This is largely sufficient to remove the homopolymer, as checked with different times of rinsing of physical mixtures of starch and pVAc. The grafting efficiency (f) is defined as:
0
10
20
30
40
50
0
20
40
60
80
100
-1.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Mono
mer
(mmo
l/min)
Temp
eratu
re (°
C)
Time (hours) Profile waterbath Pre-dosage VAM: L Pre-dosage VAM: H Pre-dosage VAM: C
Water or SPS/SBC: 10ml (10ml/min) + SPS/SBC: 30ml (5ml/hr) STS 3ml (3ml/hr)
Table 1:Taguchi L8 matrix for three factors and their interactions. AT = RPM, BT = pre-dosage of VAM, CT = pre-dosage of SPS. The applied levels (L = low, H = high and C = centre) are specified below and in Table 2.
Run Code Factors or interactionsAT BT ATBT CT ATCT BTCT ATBTCT
1 LLL -1 -1 -1 -1 -1 -1 -12 LLH -1 -1 -1 1 1 1 13 LHL -1 1 1 -1 -1 1 14 LHH -1 1 1 1 1 -1 -15 HLL 1 -1 1 -1 1 -1 16 HLH 1 -1 1 1 -1 1 -17 HHL 1 1 -1 -1 1 1 -18 HHH 1 1 -1 1 -1 -1 1
Procedure The polymerization reactor was filled with 256.2 g of demineralized water and 35.6 g of enzymatic degraded waxy potato starch. Automatic mixing (levels = 90:120:150 RPM) was started after one minute of gentle manual homogenization of the starch/water mixture in the reactor with the stirrer. The applied temperature profile and dosage protocols of VAM, SPS and STS are given in Figure 2. A total of 0.3 kg VAM was added in all cases and the levels of pre-dosage are 0, 5 and 10 % of total added VAM, respectively. The actual dosage was monitored in time and the real added amount of VAM was used for mass balance calculations. 5.3 mmol SPS and 20.0 mmol SBC were dissolved in 30-40 ml demineralised water and added in all cases as a mixture. The actual addition of the mixture starts after 104 minutes with a pre-dosage of 10 ml (10 ml/min) followed by 30 ml with a dosage speed of 5 ml/hour. The water of the pre-dosage contains 0, 12.5 or 25 wt % of the total amount of SPS and SBC solution (and therefore the remaining 30 ml contains 100, 87.5 or 75 wt % respectively). 3 ml of a 0.3 M STS solution was added with 3 ml/hour after the water bath temperature drops significantly below 65°C during the cooling down. Agitation was continued for at least one hour after the temperature of the water bath reaches a temperature of 20 °C. The dispersion was transferred into a storage container without any additional treatments after this short period of equilibration.
Figure 2:Applied temperature profile and dosage protocols of VAM, SPS/SBC and STS.
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
34 35
Table 2: Settings of the independent factors and corresponding codes.Run Code Factors
Stirring RPM
(RPM)
Pre-dosageVAM
(wt %)
Pre-dosageSPS
(wt %)1 LLL 90 0 02 LLH 90 0 25.03 LHL 90 10 04 LHH 90 10 25.05 HLL 150 0 06 HLH 150 0 25.07 HHL 150 10 08 HHH 150 10 25.09 CCC 120 5 12.5
An equilibrium HST of approximately 76 °C was observed with the reactor configuration used. A sharp decrease in the HST to a level of 66 °C was observed just after the initiation of the VAM addition. This is very close to the boiling point of the optimal composition of the binairy azeotrope water and VAM. This azeotropic refluxing theory was supported by the fact that the HST quickly increases after the VAM dosage was stopped. However, there was a considerable chance that the low boiling azeotrope of water and VAM formed was contaminated with the dissociation products of VAM (i.e. ethanal and acetic acid) or other components present in the reaction mixture. It is therefore important to realize that we are probably looking at the boiling of a contaminated binary azeotrope. The level of contamination was most likely very low due to the fact that the HST resembles the optimal boiling point of the binary azeotrope very closely. Therefore, a determination of the actual composition of the boiling azeotrope during processing was therefore not performed. The initial drop in HST coincides with a peak in WPC and these two processes can be linked to the refluxing of the azeotrope water and VAM. Heat was transferred from the reaction mixture to the cooler during this process and additional WPC was needed to compensate the loss in heat. The WPC, or level of refluxing, decreases after this peak due to a combination of conversion of VAM into oligomers and a reduction in its dosage rate.
Figure 3: HST and WPC profiles for the polymerization at CCC in triplicate.
20
40
60
80
100
0
20
40
60
80
0 1 2 3 4 5 6 7 8 9 10 11
Powe
r con
sump
tion
water
bath
(%)
Reflu
x Tem
pera
ture (
°C)
Time (Hours)
HST: CCC1 HST: CCC2 HST: CCC3 WPC: CCC1 WPC: CCC2 WPC: CCC3
VAM addition
Equation 1:
where , and denote the amount of grafted, total and homo-polymerized VAM. The residue of the acetone extraction was taken as amount of grafted copolymer as evidenced by FT-IR measurements (see below). The total pVAc was calculated on the basis of the monomer intake and conversion values. A Spectrum 2000 FT-IR from Perkin Elmer was used to characterize , and (Diamond plate; 15 scans; Resolution 4 cm-1; Interval:1 cm-1). Tg values were calculated from total and reversing heat flow curves as measured with a mDSC from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg of dispersion).
ResultsPolymerization equipment, formulation, conditions and procedureAll polymerisation reactions were carried out in a semi-batch reactor equipped with a spiral ribbon stirrer (Figure 1). SPS was used as the initiator because of its high water solubility and the fact that sodium ions cannot react with hydroxide ions (whilst its ammonium counterpart does). SBC was added to the SPS solution in an amount to ensure a pH of the reaction mixture in the range of 5 to 7. A pH below 7 is required to avoid excessive saponification of the polyvinyl acetate present and a pH above 5 is recommended to minimize the acidic breakdown of the starch part of the formulation. This was achieved by using SBC in the present study.The HST was measured and used as a measure for the reactor temperature. The two (data not shown for brevity) are indeed closely correlated in terms of their general trend as function of time. In addition, the WPC was measured to gain insights in the level of reflux. Oxygen was not removed from the initial reaction mixture and hydroquinone containing VAM was used. This is common practice in preparation on an industrial scale. However, the presence of these two inhibitors is known to reduce the efficiency of the radical formation and propagation significantly. This was compensated by increasing the amount of SPS. The effects of RPM, pre-dosage intake of VAM and SPS on process characteristics (i.e. HST and WPC) and latex composition were determined systematically (Table 2). Pre-dosage levels of VAM were set at 0, 5 and 10 wt % and based on information found in literature [37]. For SBS, pre-dosages of 0, 12.5 and 25.0 wt % were used; higher amounts were not possible as this results in pH values above 7, which was outside the desired range (vide supra).The polymerization procedure at CCC was executed in triplicate. The reproducibility of the procedure with respect to HST and WPC was excellent as revealed by their respective trends as a function of time (Figure 3).
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
34 35
Table 2: Settings of the independent factors and corresponding codes.Run Code Factors
Stirring RPM
(RPM)
Pre-dosageVAM
(wt %)
Pre-dosageSPS
(wt %)1 LLL 90 0 02 LLH 90 0 25.03 LHL 90 10 04 LHH 90 10 25.05 HLL 150 0 06 HLH 150 0 25.07 HHL 150 10 08 HHH 150 10 25.09 CCC 120 5 12.5
An equilibrium HST of approximately 76 °C was observed with the reactor configuration used. A sharp decrease in the HST to a level of 66 °C was observed just after the initiation of the VAM addition. This is very close to the boiling point of the optimal composition of the binairy azeotrope water and VAM. This azeotropic refluxing theory was supported by the fact that the HST quickly increases after the VAM dosage was stopped. However, there was a considerable chance that the low boiling azeotrope of water and VAM formed was contaminated with the dissociation products of VAM (i.e. ethanal and acetic acid) or other components present in the reaction mixture. It is therefore important to realize that we are probably looking at the boiling of a contaminated binary azeotrope. The level of contamination was most likely very low due to the fact that the HST resembles the optimal boiling point of the binary azeotrope very closely. Therefore, a determination of the actual composition of the boiling azeotrope during processing was therefore not performed. The initial drop in HST coincides with a peak in WPC and these two processes can be linked to the refluxing of the azeotrope water and VAM. Heat was transferred from the reaction mixture to the cooler during this process and additional WPC was needed to compensate the loss in heat. The WPC, or level of refluxing, decreases after this peak due to a combination of conversion of VAM into oligomers and a reduction in its dosage rate.
Figure 3: HST and WPC profiles for the polymerization at CCC in triplicate.
20
40
60
80
100
0
20
40
60
80
0 1 2 3 4 5 6 7 8 9 10 11
Powe
r con
sump
tion
water
bath
(%)
Reflu
x Tem
pera
ture (
°C)
Time (Hours)
HST: CCC1 HST: CCC2 HST: CCC3 WPC: CCC1 WPC: CCC2 WPC: CCC3
VAM addition
Equation 1:
where , and denote the amount of grafted, total and homo-polymerized VAM. The residue of the acetone extraction was taken as amount of grafted copolymer as evidenced by FT-IR measurements (see below). The total pVAc was calculated on the basis of the monomer intake and conversion values. A Spectrum 2000 FT-IR from Perkin Elmer was used to characterize , and (Diamond plate; 15 scans; Resolution 4 cm-1; Interval:1 cm-1). Tg values were calculated from total and reversing heat flow curves as measured with a mDSC from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg of dispersion).
ResultsPolymerization equipment, formulation, conditions and procedureAll polymerisation reactions were carried out in a semi-batch reactor equipped with a spiral ribbon stirrer (Figure 1). SPS was used as the initiator because of its high water solubility and the fact that sodium ions cannot react with hydroxide ions (whilst its ammonium counterpart does). SBC was added to the SPS solution in an amount to ensure a pH of the reaction mixture in the range of 5 to 7. A pH below 7 is required to avoid excessive saponification of the polyvinyl acetate present and a pH above 5 is recommended to minimize the acidic breakdown of the starch part of the formulation. This was achieved by using SBC in the present study.The HST was measured and used as a measure for the reactor temperature. The two (data not shown for brevity) are indeed closely correlated in terms of their general trend as function of time. In addition, the WPC was measured to gain insights in the level of reflux. Oxygen was not removed from the initial reaction mixture and hydroquinone containing VAM was used. This is common practice in preparation on an industrial scale. However, the presence of these two inhibitors is known to reduce the efficiency of the radical formation and propagation significantly. This was compensated by increasing the amount of SPS. The effects of RPM, pre-dosage intake of VAM and SPS on process characteristics (i.e. HST and WPC) and latex composition were determined systematically (Table 2). Pre-dosage levels of VAM were set at 0, 5 and 10 wt % and based on information found in literature [37]. For SBS, pre-dosages of 0, 12.5 and 25.0 wt % were used; higher amounts were not possible as this results in pH values above 7, which was outside the desired range (vide supra).The polymerization procedure at CCC was executed in triplicate. The reproducibility of the procedure with respect to HST and WPC was excellent as revealed by their respective trends as a function of time (Figure 3).
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
36 37
Impact of changes in polymerization procedure on process and product characteristicsA suitable reference polymerization system based on stainless steel materials was not found in literature and the results of the triplicate CCC experiments were therefore selected as a benchmark.
Reaction conditionsThe influence of RPM and pre-dosages of VAM and SPS on reaction condition-related responses are given in Table 3.
Table 3: WPC (Sum 0-7 hours) and latex composition.Code WPC Latex composition
Sum 0-7 hours(%) pH Ethanal
(mg/g)VAM
(mg/g)Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol)
LLL 16800 5.0 0.59 4.48 26.9 3.64 0.32LLH 34100 4.9 0.55 4.82 27.4 3.33 0.34LHL 66700 5.0 0.52 3.86 27.9 3.77 0.33LHH 40000 4.9 0.60 4.00 28.8 3.44 0.32HLL 33400 5.0 0.58 4.30 26.2 3.50 0.31HLH 39500 5.0 0.58 5.01 27.5 3.55 0.33HHL 63500 5.1 0.42 3.07 22.4 4.04 0.33HHH 48500 4.9 0.48 2.75 27.5 3.50 0.31CCC1 26400 5.0 0.25 1.85 25.6 3.26 0.33CCC2 36100 5.0 0.48 3.65 25.9 3.34 0.34CCC3 35600 5.1 0.47 3.56 25.7 3.25 0.34
A more detailed overview (Table A1), containing also data for dry matter, is reported in the appendix for the sake of brevity. We start by analysing the sulfur-based anion content, i.e. sulfate and thiosulfate concentration (last two columns), at the end of the polymerization. The stoichiometry of the redox equation of thiosulfate and persulfate is 2 to 1 [7]:
Reaction scheme 1:
The residual amount of STS (as measured at the end of the process and reported in Table 3) thus constitutes an indirect measure of the unreacted SPS. According to the data, the actual conversion of SPS appears to be reproducible. Moreover, the total amount of free thiosulfate and sulfate in the latexes was on average 0.34 and 3.0-3.5 mmol respectively. The fact that the latexes still contain significant amounts of thiosulfate validated the assumption that all SPS was converted. About 0.9 mmol thiosulfate was added to the latex in total and 0.56 mmol was converted at the end. This should mean that 0.28 mmol SPS was present in the reaction mixture at the moment the STS solution was dosed. A sulfate concentration of 10.5 mmol was possible based on the amount of added SPS and this means that 7.0 to 7.5 mmol sulfate was bound to the latex and/or protective colloid. However, the actual situation during processing was more complicated. The measured amount of sulfate could be the result of a number of different processes, i.e. metal ion induced dissociation, radical transfer from a sulfate ion
A typical polymerization process involves the stages initiation, propagation and termination and the observed behaviour in HST and WPC was in line with this generalized mechanism [41,60]. The formulation did not contain amphiphilic material and thus no micelles are able to withdraw VAM from the water phase. The VAM present during initiation was therefore not only available for oligomer formation but also for azeotropic refluxing (vide supra). This stage ends at the point that enough VAM oligomers of approximately 10 units are formed and aggregate into particles [60,61]. The VAM dosage can be increased during the propagation stage because a part of the VAM can be stored in the particles present and the storage capacity increases with increasing particle size [38,39]. These considerations are of significant use from a practical point of view since they clearly indicate the HST as a significant response that can be related to the different polymerization stages. This parameter could therefore, in principle, be used to automatically adjust the monomer concentration during the polymerization process. This is confirmed by the impact of the VAM pre-dosages (runs LLL and LHL, Figure 4) on HST and WPC.Both responses were influenced in a way that exceeds the variation observed between the replicates of Figure 3, with the WPC showing more pronounced differences. This was probably related to the fact that the heat transfer in the situation with the highest VAM pre-dosage turned out to be very effective. The water bath of the reflux cooler was not able to keep the temperature at 16 °C and even temperatures up to 45 °C were observed during severe refluxing conditions. It is worth noting that this effect does not occur in reactors made of borosilicate in the same degree due the relatively higher (close to 10 times) thermal conductivity of stainless steel [16,17]. The results based on stainless steel and borosilicate based polymerization systems are therefore not interchangeable due not only to a (slight) difference in iron ions content, but a heat balance point of view as well. This simple conclusion is often, to the best of our knowledge, not mentioned in the open literature. However, it must be taken into account when trying to extrapolate reaction mechanisms and process conditions from lab-scale (glass) equipment to industrial (steel) one.
Figure 4: HST and WPC profiles for the polymerization at CCC compared to two counterparts with a high (LHL) and low (LLL) pre-dosage of VAM.
20
40
60
80
100
0
20
40
60
80
0 1 2 3 4 5 6 7 8 9 10 11
Powe
r Wate
rbath
(%)
Reflu
x Tem
pera
ture (
°C)
Time (Hours)
HST: Average(CCC) HST: LLL HST: LHL WPC: Average(CCC) WPC: LLL WPC: LHL
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
36 37
Impact of changes in polymerization procedure on process and product characteristicsA suitable reference polymerization system based on stainless steel materials was not found in literature and the results of the triplicate CCC experiments were therefore selected as a benchmark.
Reaction conditionsThe influence of RPM and pre-dosages of VAM and SPS on reaction condition-related responses are given in Table 3.
Table 3: WPC (Sum 0-7 hours) and latex composition.Code WPC Latex composition
Sum 0-7 hours(%) pH Ethanal
(mg/g)VAM
(mg/g)Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol)
LLL 16800 5.0 0.59 4.48 26.9 3.64 0.32LLH 34100 4.9 0.55 4.82 27.4 3.33 0.34LHL 66700 5.0 0.52 3.86 27.9 3.77 0.33LHH 40000 4.9 0.60 4.00 28.8 3.44 0.32HLL 33400 5.0 0.58 4.30 26.2 3.50 0.31HLH 39500 5.0 0.58 5.01 27.5 3.55 0.33HHL 63500 5.1 0.42 3.07 22.4 4.04 0.33HHH 48500 4.9 0.48 2.75 27.5 3.50 0.31CCC1 26400 5.0 0.25 1.85 25.6 3.26 0.33CCC2 36100 5.0 0.48 3.65 25.9 3.34 0.34CCC3 35600 5.1 0.47 3.56 25.7 3.25 0.34
A more detailed overview (Table A1), containing also data for dry matter, is reported in the appendix for the sake of brevity. We start by analysing the sulfur-based anion content, i.e. sulfate and thiosulfate concentration (last two columns), at the end of the polymerization. The stoichiometry of the redox equation of thiosulfate and persulfate is 2 to 1 [7]:
Reaction scheme 1:
The residual amount of STS (as measured at the end of the process and reported in Table 3) thus constitutes an indirect measure of the unreacted SPS. According to the data, the actual conversion of SPS appears to be reproducible. Moreover, the total amount of free thiosulfate and sulfate in the latexes was on average 0.34 and 3.0-3.5 mmol respectively. The fact that the latexes still contain significant amounts of thiosulfate validated the assumption that all SPS was converted. About 0.9 mmol thiosulfate was added to the latex in total and 0.56 mmol was converted at the end. This should mean that 0.28 mmol SPS was present in the reaction mixture at the moment the STS solution was dosed. A sulfate concentration of 10.5 mmol was possible based on the amount of added SPS and this means that 7.0 to 7.5 mmol sulfate was bound to the latex and/or protective colloid. However, the actual situation during processing was more complicated. The measured amount of sulfate could be the result of a number of different processes, i.e. metal ion induced dissociation, radical transfer from a sulfate ion
A typical polymerization process involves the stages initiation, propagation and termination and the observed behaviour in HST and WPC was in line with this generalized mechanism [41,60]. The formulation did not contain amphiphilic material and thus no micelles are able to withdraw VAM from the water phase. The VAM present during initiation was therefore not only available for oligomer formation but also for azeotropic refluxing (vide supra). This stage ends at the point that enough VAM oligomers of approximately 10 units are formed and aggregate into particles [60,61]. The VAM dosage can be increased during the propagation stage because a part of the VAM can be stored in the particles present and the storage capacity increases with increasing particle size [38,39]. These considerations are of significant use from a practical point of view since they clearly indicate the HST as a significant response that can be related to the different polymerization stages. This parameter could therefore, in principle, be used to automatically adjust the monomer concentration during the polymerization process. This is confirmed by the impact of the VAM pre-dosages (runs LLL and LHL, Figure 4) on HST and WPC.Both responses were influenced in a way that exceeds the variation observed between the replicates of Figure 3, with the WPC showing more pronounced differences. This was probably related to the fact that the heat transfer in the situation with the highest VAM pre-dosage turned out to be very effective. The water bath of the reflux cooler was not able to keep the temperature at 16 °C and even temperatures up to 45 °C were observed during severe refluxing conditions. It is worth noting that this effect does not occur in reactors made of borosilicate in the same degree due the relatively higher (close to 10 times) thermal conductivity of stainless steel [16,17]. The results based on stainless steel and borosilicate based polymerization systems are therefore not interchangeable due not only to a (slight) difference in iron ions content, but a heat balance point of view as well. This simple conclusion is often, to the best of our knowledge, not mentioned in the open literature. However, it must be taken into account when trying to extrapolate reaction mechanisms and process conditions from lab-scale (glass) equipment to industrial (steel) one.
Figure 4: HST and WPC profiles for the polymerization at CCC compared to two counterparts with a high (LHL) and low (LLL) pre-dosage of VAM.
20
40
60
80
100
0
20
40
60
80
0 1 2 3 4 5 6 7 8 9 10 11
Powe
r Wate
rbath
(%)
Reflu
x Tem
pera
ture (
°C)
Time (Hours)
HST: Average(CCC) HST: LLL HST: LHL WPC: Average(CCC) WPC: LLL WPC: LHL
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
38 39
the added SPS solution was around 8 and products prepared with a pre-dosage of SPS therefore, had a somewhat higher initial pH than their counterparts without this pre-dosage. The concentration of acetate ions decreased with increasing RPM and this might be explained by the correlation between RPM and degree of mixing. The life time of small drops of alkaline SPS solution was probably much shorter at high RPM and this could explain the lower degree of saponification of VAM and/or pVAc.
Latex characteristicsThe applied variations in RPM, VAM and SPS have a distinct impact (a complete overview is provided in Table A2, Appendix) on the viscosity and the PSD of the latexes after reaction. The observed variations in the stability response were comparable in all measurements and were indicative for sufficiently stable dispersions during measurement at least. The significant impact of the settings on the process characteristics was statistically confirmed (Table 5) by the analysis of the corresponding effects, which indicate that VAM pre-dosage variations were largely responsible for the observed viscosity and particle size effects.
Table 5: Effect of different reaction conditions on the viscosity and PSD (Table A2) based on Taguchi calculations (simplified). Only effects are shown which exceed two times the σ of the triplicate of the centre settings.
Factors Viscosity PSDPeak Particles with diameter <1 mm
(mPa·s)Mode(mm)
Half Width(mm)
Amount(%)
Stability(Max-Min)
RPM - - -0.99 9 -VAM -78 -0.39 -2.29 15 -RPM:VAM - - -1.10 - -SPS - - - - -SPS:RPM - - - - -SPS:VAM - - - - -SPS:RPM:VAM - - - - -CCC: Average 118 0.85 5.07 61 0.2CCC: 2*σ 21 0.19 0.72 8 0.2
The mode and the half width of the PSD were both affected by changes in the VAM pre-dosage, which suggested that the half width was linked not only to the initial particle formation stage, but also to secondary particle formation. RPM had a strong effect, by itself and in combination with a pre-dosage of VAM, on the half width of the distribution. The fact that the mode of the distribution did not change, by this factor and its interaction, suggest that this variable was connected to particle formation during the actual polymerization mainly. This tentative explanation is corroborated by the fact that differences in the level of agitation will affect the chance of particles meeting active oligomeric radicals formed and therefore influence the chance of additional particle formation during polymerization [35]. Independently of the exact mechanism (beyond the scope of this work), it must be stressed that the VAM pre-dosage levels play an important role. In particular, a high VAM pre-dosage appears to be a necessity to generate a PSD between 0.5 and 32.5 μm (Figure 5 & 6).
to another compound in the formulation or dissociation with and without radical formation [7,62,63]. Besides the important indication of full SPS conversion, no other conclusion can be drawn and further research is in this case needed to fully explain the observed differences in anion composition properly.The variation in reaction conditions had a distinct influence on the WPC during processing but the impact on VAM conversion, radical formation mechanism (pH, sulfate & thiosulfate) and composition (recovery, ethanal & acetate) appeared to be limited. The simplified Taguchi procedure was applied on the results of Table 3. The effects of the three selected reaction condition related factors and their interactions, which are larger than two times the standard deviation (σ) of the centre settings, are shown in Table 4. According the simplified Taguchi calculations, WPC and concentrations of acetate and sulfate can be linked in a statistically significant way to the independent variables. The sign of the statistical coefficients indicate a positive correlation of WPC with VAM and SPS*VAM. The increased level of WPC after a pre-dosage of VAM was in line with the expectations, because the level of azeotropic refluxing depends strongly on the amount of excess VAM in the reaction mixture. The observed interaction between the pre-dosages VAM and SPS was also expected, because radical formation of SPS occurs by thermal dissociation and the temperature of the reaction mixture was sensitive for changes in level of reflux (section 3.1).
Table 4: Effect of different reaction conditions on the 7 selected variables based on Taguchi calculations (simplified). Only effects are shown which exceed the threshold of two times the standard deviation (σ) of the triplicate of the centre settings.
Factors WPC Latex compositionSum 0-7
hours(%) pH
Ethanal(mg/g)
VAM(mmol)
Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol)
RPM - - - - -1.82 0.11 -VAM 23700 - - - -0.33 0.18 -RPM:VAM - - - - 1.56 - -SPS - -0.10 - - 1.96 -0.28 -SPS:RPM - - - - -1.25 - -SPS:VAM 16300 0.05 - - -1.08 0.15 0.02SPS:RPM:VAM - - - - 0.87 -0.14 -CCC: Average 32700 5.04 0.40 3.00 25.70 3.28 0.34CCCC: 2*σ 10900 0.05 0.30 2.00 0.27 0.09 0.02
The amount of free acetate present at the end of the polymerization was basically a function of all investigated factors, particularly RPM and SPS intake (Table 4). Also in this case the observed trend can be rationalized and linked to the employed reaction conditions. Acetate ions are not only generated by saponification during processing but also during the pre-treatment needed for the anion determination. The solution of the latex still contained short polyvinyl acetate chains after the centrifugation step and these remnants saponify as soon as the excess sodium hydroxide was added. The part of acetate ions which originate from this process was, however, assumed to be similar in all executed experiments. The pH of
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
38 39
the added SPS solution was around 8 and products prepared with a pre-dosage of SPS therefore, had a somewhat higher initial pH than their counterparts without this pre-dosage. The concentration of acetate ions decreased with increasing RPM and this might be explained by the correlation between RPM and degree of mixing. The life time of small drops of alkaline SPS solution was probably much shorter at high RPM and this could explain the lower degree of saponification of VAM and/or pVAc.
Latex characteristicsThe applied variations in RPM, VAM and SPS have a distinct impact (a complete overview is provided in Table A2, Appendix) on the viscosity and the PSD of the latexes after reaction. The observed variations in the stability response were comparable in all measurements and were indicative for sufficiently stable dispersions during measurement at least. The significant impact of the settings on the process characteristics was statistically confirmed (Table 5) by the analysis of the corresponding effects, which indicate that VAM pre-dosage variations were largely responsible for the observed viscosity and particle size effects.
Table 5: Effect of different reaction conditions on the viscosity and PSD (Table A2) based on Taguchi calculations (simplified). Only effects are shown which exceed two times the σ of the triplicate of the centre settings.
Factors Viscosity PSDPeak Particles with diameter <1 mm
(mPa·s)Mode(mm)
Half Width(mm)
Amount(%)
Stability(Max-Min)
RPM - - -0.99 9 -VAM -78 -0.39 -2.29 15 -RPM:VAM - - -1.10 - -SPS - - - - -SPS:RPM - - - - -SPS:VAM - - - - -SPS:RPM:VAM - - - - -CCC: Average 118 0.85 5.07 61 0.2CCC: 2*σ 21 0.19 0.72 8 0.2
The mode and the half width of the PSD were both affected by changes in the VAM pre-dosage, which suggested that the half width was linked not only to the initial particle formation stage, but also to secondary particle formation. RPM had a strong effect, by itself and in combination with a pre-dosage of VAM, on the half width of the distribution. The fact that the mode of the distribution did not change, by this factor and its interaction, suggest that this variable was connected to particle formation during the actual polymerization mainly. This tentative explanation is corroborated by the fact that differences in the level of agitation will affect the chance of particles meeting active oligomeric radicals formed and therefore influence the chance of additional particle formation during polymerization [35]. Independently of the exact mechanism (beyond the scope of this work), it must be stressed that the VAM pre-dosage levels play an important role. In particular, a high VAM pre-dosage appears to be a necessity to generate a PSD between 0.5 and 32.5 μm (Figure 5 & 6).
to another compound in the formulation or dissociation with and without radical formation [7,62,63]. Besides the important indication of full SPS conversion, no other conclusion can be drawn and further research is in this case needed to fully explain the observed differences in anion composition properly.The variation in reaction conditions had a distinct influence on the WPC during processing but the impact on VAM conversion, radical formation mechanism (pH, sulfate & thiosulfate) and composition (recovery, ethanal & acetate) appeared to be limited. The simplified Taguchi procedure was applied on the results of Table 3. The effects of the three selected reaction condition related factors and their interactions, which are larger than two times the standard deviation (σ) of the centre settings, are shown in Table 4. According the simplified Taguchi calculations, WPC and concentrations of acetate and sulfate can be linked in a statistically significant way to the independent variables. The sign of the statistical coefficients indicate a positive correlation of WPC with VAM and SPS*VAM. The increased level of WPC after a pre-dosage of VAM was in line with the expectations, because the level of azeotropic refluxing depends strongly on the amount of excess VAM in the reaction mixture. The observed interaction between the pre-dosages VAM and SPS was also expected, because radical formation of SPS occurs by thermal dissociation and the temperature of the reaction mixture was sensitive for changes in level of reflux (section 3.1).
Table 4: Effect of different reaction conditions on the 7 selected variables based on Taguchi calculations (simplified). Only effects are shown which exceed the threshold of two times the standard deviation (σ) of the triplicate of the centre settings.
Factors WPC Latex compositionSum 0-7
hours(%) pH
Ethanal(mg/g)
VAM(mmol)
Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol)
RPM - - - - -1.82 0.11 -VAM 23700 - - - -0.33 0.18 -RPM:VAM - - - - 1.56 - -SPS - -0.10 - - 1.96 -0.28 -SPS:RPM - - - - -1.25 - -SPS:VAM 16300 0.05 - - -1.08 0.15 0.02SPS:RPM:VAM - - - - 0.87 -0.14 -CCC: Average 32700 5.04 0.40 3.00 25.70 3.28 0.34CCCC: 2*σ 10900 0.05 0.30 2.00 0.27 0.09 0.02
The amount of free acetate present at the end of the polymerization was basically a function of all investigated factors, particularly RPM and SPS intake (Table 4). Also in this case the observed trend can be rationalized and linked to the employed reaction conditions. Acetate ions are not only generated by saponification during processing but also during the pre-treatment needed for the anion determination. The solution of the latex still contained short polyvinyl acetate chains after the centrifugation step and these remnants saponify as soon as the excess sodium hydroxide was added. The part of acetate ions which originate from this process was, however, assumed to be similar in all executed experiments. The pH of
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
40 41
starch containing material [64]. Acetone is a good solvent for pVAc also and is recommended to be used instead of benzene from an environmental and health point of view [65]. A cold acetone precipitation step was added to the described procedure to remove excess pVAc homopolymer. The procedure resulted in a residue and an acetone soluble fraction. Examples of typical FT-IR spectra’s of these two materials are given in Figure 7. The material in both acetone fractions (i.e. precipitation and extraction, here taken together) appear to be pVAc only (see for example –CO- stretching vibration at 1730 cm-1) and the residue after the soxhlet extraction shoed signs of both pVAc and starch (besides the peak of pVAc outlined above, see the –OH stretching vibration at n > 3000 cm-1) [64,66].
Figure 7: FT-IR spectra of residue (below) and acetone soluble fraction (above).
f (as defined in the experimental part) was calculated on the basis of the amount of the grafted VAM (Table 6). Already at first glance it seems that the initiator pre-dosage (the third H or L in the samples code) was the decisive factor in determining the f values. In particular the SPS pre-dosage resulted in almost no grafting formation (f ≈ 0, see entries 2, 4, 6 and 8 in Table 6). This is confirmed by the Taguchi calculations (Table 7) of the selected reaction conditions and their interactions.Variations in VAM and SPS dosages have shown the most pronounced effects (among all studied factors and interactions) on the grafting efficiency, the latter increasing with VAM pre-dosage and decreasing (largely) with the SPS one. Both pre-dosages had a similar effect on the free sulfate concentration of the latex (see Table 3). This might be explained by a more prominent radical transfer from sulfate to starch (in case of SPS pre-dosage) and thus by the occurrence of side reactions (e.g. due to cage effects). The application of a VAM pre-dosage resulted in an increased grafting efficiency and improved the monodispersity of the latex (Table 5). These effects are usually achieved by micelle generation during the initial stage of the polymerization by adding sufficiently amphiphilic ingredients to the formulation [37].
0.0
0.5
1.0
1.5
2.0
500 1000 1500 2000 2500 3000 3500 4000
Abso
rban
ce un
its
Wave number (cm-1)
Figure 5: PSD’s of the latexes prepared at low RPM (compared to the averaged PSD of CCC).
Figure 6: PSD’s of the latexes prepared at high RPM (compared to the averaged PSD of CCC).
Protective colloid related propertiesThe applied reaction conditions resulted in a number of interesting changes in the PSD’s. These may be related to the amount of starch with pVAc grafts generated during the polymerization. Exhaustive soxhlet extractions with benzene are suitable to separate pVAc homopolymer from
0.0
0.5
1.0
1.5
2.0
2.5
0.1 1 10 100
Dens
ity
Size (µm)
Average(CCC) HLL HLH HHL HHH
0.0
0.5
1.0
1.5
2.0
2.5
0.1 1 10 100
Dens
ity
Size (µm)
Average(CCC) LLL LLH LHL LHH
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
40 41
starch containing material [64]. Acetone is a good solvent for pVAc also and is recommended to be used instead of benzene from an environmental and health point of view [65]. A cold acetone precipitation step was added to the described procedure to remove excess pVAc homopolymer. The procedure resulted in a residue and an acetone soluble fraction. Examples of typical FT-IR spectra’s of these two materials are given in Figure 7. The material in both acetone fractions (i.e. precipitation and extraction, here taken together) appear to be pVAc only (see for example –CO- stretching vibration at 1730 cm-1) and the residue after the soxhlet extraction shoed signs of both pVAc and starch (besides the peak of pVAc outlined above, see the –OH stretching vibration at n > 3000 cm-1) [64,66].
Figure 7: FT-IR spectra of residue (below) and acetone soluble fraction (above).
f (as defined in the experimental part) was calculated on the basis of the amount of the grafted VAM (Table 6). Already at first glance it seems that the initiator pre-dosage (the third H or L in the samples code) was the decisive factor in determining the f values. In particular the SPS pre-dosage resulted in almost no grafting formation (f ≈ 0, see entries 2, 4, 6 and 8 in Table 6). This is confirmed by the Taguchi calculations (Table 7) of the selected reaction conditions and their interactions.Variations in VAM and SPS dosages have shown the most pronounced effects (among all studied factors and interactions) on the grafting efficiency, the latter increasing with VAM pre-dosage and decreasing (largely) with the SPS one. Both pre-dosages had a similar effect on the free sulfate concentration of the latex (see Table 3). This might be explained by a more prominent radical transfer from sulfate to starch (in case of SPS pre-dosage) and thus by the occurrence of side reactions (e.g. due to cage effects). The application of a VAM pre-dosage resulted in an increased grafting efficiency and improved the monodispersity of the latex (Table 5). These effects are usually achieved by micelle generation during the initial stage of the polymerization by adding sufficiently amphiphilic ingredients to the formulation [37].
0.0
0.5
1.0
1.5
2.0
500 1000 1500 2000 2500 3000 3500 4000
Abso
rban
ce un
its
Wave number (cm-1)
Figure 5: PSD’s of the latexes prepared at low RPM (compared to the averaged PSD of CCC).
Figure 6: PSD’s of the latexes prepared at high RPM (compared to the averaged PSD of CCC).
Protective colloid related propertiesThe applied reaction conditions resulted in a number of interesting changes in the PSD’s. These may be related to the amount of starch with pVAc grafts generated during the polymerization. Exhaustive soxhlet extractions with benzene are suitable to separate pVAc homopolymer from
0.0
0.5
1.0
1.5
2.0
2.5
0.1 1 10 100
Dens
ity
Size (µm)
Average(CCC) HLL HLH HHL HHH
0.0
0.5
1.0
1.5
2.0
2.5
0.1 1 10 100
Dens
ity
Size (µm)
Average(CCC) LLL LLH LHL LHH
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
42 43
flow diagram was predominantly influenced by changes in pre-dosages of VAM, SPS and their interaction. The f is influenced by these two variables in a similar way (Table 7) and this makes a certain correlation between Tg,inflection and degree of grafting plausible. An increase in f is expected to result in lower Tg,inflection values due to an increased level of heterogeneity. However, this is not in line with the sign of the corresponding coefficient (+1.0 in Table 9). On the other hand, a change in Tg,inflection can be also related to a different polymer configuration, which might be relevant in this case since the average chain length and level of branching of pVAc are known to be temperature dependent [38,67,68]. A pre-dosage of VAM results in a higher Tg,inflection and this an indication for a longer average chain length, a lower degree of branching or a combination of both. Chain length and branching are both correlated to the reaction temperature and the factor VAM and the interaction VAM*SPS should coincide with a lower reaction temperature if this assumption is valid. This is probably true based on the results of the Taguchi calculations of the response WPC (Table 3). A high level of WPC was not only observed for this factor but also for the interaction. A high WPC corresponds with a high level of reflux and this process results in a reaction temperature close to the boiling point of the azeotrope VAM/water (66°C) (section 3.1). A lower reaction temperature also diminishes the amount of radicals formed because SPS was used in the thermal dissociation mode. A change in the amount of radicals probably influenced the number of pVAc chains and branches formed during polymerization and this process might contribute to the differences in polymer configuration introduced by the factor VAM and the interaction of VAM*SPS as well. The assumption that the number of radicals formed was linked to the level of change in polymer configuration is supported by the effect of SPS on the Tg,inflecion. A pre-dosage of SPS resulted in a considerable lower Tg,inflection.
Figure 8: An example of a typical obtained heat flow diagram. The non-reversing heat flow is shown as a solid line and the total and reversing heat flows are represented by a cross and plus respectively. The calculation is also visually depicted together with the placement of the thermal transitions onset, inflection and endset. The glass transition temperature is in this research defined as the inflection point of the transition (Tg,inflection).
0.00
0.04
0.08
0.12
0.16
0.20
0.7
0.8
0.9
1.0
1.1
1.2
0 5 10 15 20 25
Rev H
eat F
low (W
/g)
Heat
Flow
(W/g)
Temperature (°C)
Inflection
Onset
Endothermic volume relaxation
Inflection Endset
Endset
Onset
Table 6: Amount of homopolymer, grafted polymer and f for all prepared latexes.Code pVAc composition
Homopolymer(mg/g)
Grafted polymer(mg/g)
f(fraction)
LLL 389 79 0.17LLH 464 5 0.01LHL 368 98 0.21LHH 454 12 0.03HLL 391 77 0.16HLH 465 1 0.00HHL 279 191 0.41HHH 461 9 0.02CCC1 371 92 0.20CCC2 376 91 0.19CCC3 378 85 0.18
Table 7: Effect of reaction conditions on the selected pVAc composition variables (based on simplified Taguchi calculations). Only effects are shown which exceed two times the σ of the triplicate of the centre settings.
Factors pVAc compositionHomopolymer
(mg/g)Grafted polymer
(mg/g)f
(fraction)RPM -20 21 0.04VAM -37 37 0.08RPM:VAM 21 -24 -0.05SPS 104 -104 -0.22SPS:RPM -24 24 0.05SPS:VAM -30 29 0.06SPS:RPM:VAM 24 -23 -0.05CCC: Average 375 89 0.19CCC: 2*σ 7 7 0.02
Polyvinyl acetate characteristicsGlass transition temperatures onset (Tg,onset), inflection (Tg,inflection), and endset (Tg,endset) (for definition see Figure 8), of the latexes were determined by mDSC (for details see Table A4, Appendix). The simplified Taguchi calculations are shown in Table 8. The Tg,inflection increases with approximately 3°C if the calculations are based on the reversing heat flow instead of its total counterpart. The removal of the endothermic volume relaxation components (Figure 8) did not only lead to a higher Tg,inflection but to a considerable broader glass transition trajectory as well. Fortunately, the increased trajectory coincides with an improved level of reproducibility according the triplicate measurement of CCC. The actual position of the Tg,inflection in the heat
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
42 43
flow diagram was predominantly influenced by changes in pre-dosages of VAM, SPS and their interaction. The f is influenced by these two variables in a similar way (Table 7) and this makes a certain correlation between Tg,inflection and degree of grafting plausible. An increase in f is expected to result in lower Tg,inflection values due to an increased level of heterogeneity. However, this is not in line with the sign of the corresponding coefficient (+1.0 in Table 9). On the other hand, a change in Tg,inflection can be also related to a different polymer configuration, which might be relevant in this case since the average chain length and level of branching of pVAc are known to be temperature dependent [38,67,68]. A pre-dosage of VAM results in a higher Tg,inflection and this an indication for a longer average chain length, a lower degree of branching or a combination of both. Chain length and branching are both correlated to the reaction temperature and the factor VAM and the interaction VAM*SPS should coincide with a lower reaction temperature if this assumption is valid. This is probably true based on the results of the Taguchi calculations of the response WPC (Table 3). A high level of WPC was not only observed for this factor but also for the interaction. A high WPC corresponds with a high level of reflux and this process results in a reaction temperature close to the boiling point of the azeotrope VAM/water (66°C) (section 3.1). A lower reaction temperature also diminishes the amount of radicals formed because SPS was used in the thermal dissociation mode. A change in the amount of radicals probably influenced the number of pVAc chains and branches formed during polymerization and this process might contribute to the differences in polymer configuration introduced by the factor VAM and the interaction of VAM*SPS as well. The assumption that the number of radicals formed was linked to the level of change in polymer configuration is supported by the effect of SPS on the Tg,inflecion. A pre-dosage of SPS resulted in a considerable lower Tg,inflection.
Figure 8: An example of a typical obtained heat flow diagram. The non-reversing heat flow is shown as a solid line and the total and reversing heat flows are represented by a cross and plus respectively. The calculation is also visually depicted together with the placement of the thermal transitions onset, inflection and endset. The glass transition temperature is in this research defined as the inflection point of the transition (Tg,inflection).
0.00
0.04
0.08
0.12
0.16
0.20
0.7
0.8
0.9
1.0
1.1
1.2
0 5 10 15 20 25
Rev H
eat F
low (W
/g)
Heat
Flow
(W/g)
Temperature (°C)
Inflection
Onset
Endothermic volume relaxation
Inflection Endset
Endset
Onset
Table 6: Amount of homopolymer, grafted polymer and f for all prepared latexes.Code pVAc composition
Homopolymer(mg/g)
Grafted polymer(mg/g)
f(fraction)
LLL 389 79 0.17LLH 464 5 0.01LHL 368 98 0.21LHH 454 12 0.03HLL 391 77 0.16HLH 465 1 0.00HHL 279 191 0.41HHH 461 9 0.02CCC1 371 92 0.20CCC2 376 91 0.19CCC3 378 85 0.18
Table 7: Effect of reaction conditions on the selected pVAc composition variables (based on simplified Taguchi calculations). Only effects are shown which exceed two times the σ of the triplicate of the centre settings.
Factors pVAc compositionHomopolymer
(mg/g)Grafted polymer
(mg/g)f
(fraction)RPM -20 21 0.04VAM -37 37 0.08RPM:VAM 21 -24 -0.05SPS 104 -104 -0.22SPS:RPM -24 24 0.05SPS:VAM -30 29 0.06SPS:RPM:VAM 24 -23 -0.05CCC: Average 375 89 0.19CCC: 2*σ 7 7 0.02
Polyvinyl acetate characteristicsGlass transition temperatures onset (Tg,onset), inflection (Tg,inflection), and endset (Tg,endset) (for definition see Figure 8), of the latexes were determined by mDSC (for details see Table A4, Appendix). The simplified Taguchi calculations are shown in Table 8. The Tg,inflection increases with approximately 3°C if the calculations are based on the reversing heat flow instead of its total counterpart. The removal of the endothermic volume relaxation components (Figure 8) did not only lead to a higher Tg,inflection but to a considerable broader glass transition trajectory as well. Fortunately, the increased trajectory coincides with an improved level of reproducibility according the triplicate measurement of CCC. The actual position of the Tg,inflection in the heat
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
44 45
two distinctive aspects. In first place the use of experimental conditions closely resembling industrial settings allows the establishment of a reliable process-product relationship to be used beyond laboratory scales. In this context, the observed effects of the experimental conditions on the process and product characteristics have been tentatively explained on the basis of the available literature on the topic as well as the peculiar settings employed here. Secondly, the statistical analysis factually constitutes a “theoretical” toolbox never explored for this polymerization reaction under the employed reaction conditions before. Besides, its inherent simplicity, constitutes a relevant step forward towards the (industrial) implementation of modified starch as protective colloids in radical polymerization reactions. As such it constitutes the starting point for further optimization of the process under consideration.
AcknowledgementsThis investigation was sponsored by Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
AbbreviationspVAc : Polyvinyl acetate-based polymer.VAM : Vinyl acetate monomer.pVOH : Polyvinyl alcohol.PSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulated diffential scanning calorimeter.SPS : Sodium persulfate.SBC : Sodium bicarbonate.STS : Sodium thiosulfate.HST : Headspace temperature.WPC : Water bath power consumption.RPM : Revolutions per minute of stirrer.AT : RPM.BT : Pre-dosage of VAM.CT : Pre-dosage of initiator/buffer (SPS).L : Low level.H : High level.C : Centre level.f : Grafting efficiency.
: Amount of grafted pVAc : Amount of total pVAc : Amount of homopolymer pVAc.σ : Standard deviation.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
Table 8: Effect of different reaction conditions on the selected product characterization responses (Table A4, Appendix) based on Taguchi calculations (simplified). Only effects are shown which exceed two times the σ of the triplicate of the centre settings.
Factors Thermal transitionsTotal heat flow Reversing heat flow
Tg,inflection DTg Tg,inflection DTg(°C) (°C) (°C) (°C)
RPM - - - -VAM 1.3 - 1.0 -RPM:VAM - - - -SPS - 0.9 - - 1.4 -SPS:RPM - - - -SPS:VAM 0.7 - 1.1 -SPS:RPM:VAM - - - -CCC: Average 9.9 3.1 13.1 4.4CCC: 2*σ 0.7 1.6 0.7 0.9
ConclusionsThe designed polymerization system is suitable to prepare modified starch stabilized vinyl acetate dispersions in a temperature range of 65 - 85 °C in a reproducible way. The setup allowed the application of distinct different levels of agitation and pre-dosages of initiator/buffer and vinyl acetate monomer. The selected monitoring and latex characterization techniques captured the influence of these responses on important aspects of reaction conditions and properties of the final product. Degrees of conversion and dry matter content of the obtained dispersions were comparable and allowed a proper evaluation of the effects of the applied reaction conditions on the selected latex related responses.The applied levels of agitation and selected pre-dosage levels of monomer significantly influenced the particle size distribution of the dispersion but only to a limited extend. The selected pre-dosages of monomer and initiator/buffer showed a clear but modest influence on the glass transition temperature of the generated polyvinyl acetate polymer. The amount of monomer pre-dosage had a considerable effect on the viscosity of the latex. However, a high pre-dosage of monomer also resulted in heat transfer from the reaction mixture to the reflux cooler, which was barely acceptable with respect to a proper temperature control and homogeneity of the reaction mixture during polymerization. The selected reaction conditions did result in considerable changes in level of protective colloid grafted (and/or complex mixture material) to the particles present in the latex. Furthermore, significant changes in pH and anion composition were observed but only to a modest degree. The results can be used to evaluate the impact of the reaction conditions on the actual radical formation process, but the current experimental set-up does not allow the definition of solid conclusions in this area yet. Further research is required to address this important aspect of the polymerization properly. However, the observed effects clearly stress the need of a suitable reference system (e.g. in connection of the reactor type and material) when trying to make allowances for the applicability at industrial level. The obtained results constitute in our opinion to a significant novelty in open literature for
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
44 45
two distinctive aspects. In first place the use of experimental conditions closely resembling industrial settings allows the establishment of a reliable process-product relationship to be used beyond laboratory scales. In this context, the observed effects of the experimental conditions on the process and product characteristics have been tentatively explained on the basis of the available literature on the topic as well as the peculiar settings employed here. Secondly, the statistical analysis factually constitutes a “theoretical” toolbox never explored for this polymerization reaction under the employed reaction conditions before. Besides, its inherent simplicity, constitutes a relevant step forward towards the (industrial) implementation of modified starch as protective colloids in radical polymerization reactions. As such it constitutes the starting point for further optimization of the process under consideration.
AcknowledgementsThis investigation was sponsored by Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
AbbreviationspVAc : Polyvinyl acetate-based polymer.VAM : Vinyl acetate monomer.pVOH : Polyvinyl alcohol.PSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulated diffential scanning calorimeter.SPS : Sodium persulfate.SBC : Sodium bicarbonate.STS : Sodium thiosulfate.HST : Headspace temperature.WPC : Water bath power consumption.RPM : Revolutions per minute of stirrer.AT : RPM.BT : Pre-dosage of VAM.CT : Pre-dosage of initiator/buffer (SPS).L : Low level.H : High level.C : Centre level.f : Grafting efficiency.
: Amount of grafted pVAc : Amount of total pVAc : Amount of homopolymer pVAc.σ : Standard deviation.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
Table 8: Effect of different reaction conditions on the selected product characterization responses (Table A4, Appendix) based on Taguchi calculations (simplified). Only effects are shown which exceed two times the σ of the triplicate of the centre settings.
Factors Thermal transitionsTotal heat flow Reversing heat flow
Tg,inflection DTg Tg,inflection DTg(°C) (°C) (°C) (°C)
RPM - - - -VAM 1.3 - 1.0 -RPM:VAM - - - -SPS - 0.9 - - 1.4 -SPS:RPM - - - -SPS:VAM 0.7 - 1.1 -SPS:RPM:VAM - - - -CCC: Average 9.9 3.1 13.1 4.4CCC: 2*σ 0.7 1.6 0.7 0.9
ConclusionsThe designed polymerization system is suitable to prepare modified starch stabilized vinyl acetate dispersions in a temperature range of 65 - 85 °C in a reproducible way. The setup allowed the application of distinct different levels of agitation and pre-dosages of initiator/buffer and vinyl acetate monomer. The selected monitoring and latex characterization techniques captured the influence of these responses on important aspects of reaction conditions and properties of the final product. Degrees of conversion and dry matter content of the obtained dispersions were comparable and allowed a proper evaluation of the effects of the applied reaction conditions on the selected latex related responses.The applied levels of agitation and selected pre-dosage levels of monomer significantly influenced the particle size distribution of the dispersion but only to a limited extend. The selected pre-dosages of monomer and initiator/buffer showed a clear but modest influence on the glass transition temperature of the generated polyvinyl acetate polymer. The amount of monomer pre-dosage had a considerable effect on the viscosity of the latex. However, a high pre-dosage of monomer also resulted in heat transfer from the reaction mixture to the reflux cooler, which was barely acceptable with respect to a proper temperature control and homogeneity of the reaction mixture during polymerization. The selected reaction conditions did result in considerable changes in level of protective colloid grafted (and/or complex mixture material) to the particles present in the latex. Furthermore, significant changes in pH and anion composition were observed but only to a modest degree. The results can be used to evaluate the impact of the reaction conditions on the actual radical formation process, but the current experimental set-up does not allow the definition of solid conclusions in this area yet. Further research is required to address this important aspect of the polymerization properly. However, the observed effects clearly stress the need of a suitable reference system (e.g. in connection of the reactor type and material) when trying to make allowances for the applicability at industrial level. The obtained results constitute in our opinion to a significant novelty in open literature for
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
46 47
factors in emulsion polymerization process development, Polym. React. Eng. 11 (3) (2003) 259-276.[26] M. Nomura, M. Harada, W. Eguchi, S. Nagata, Effect of stirring on the emulsion polymerization of styrene,
J. Appl. Polym. Sci. 16 (1972) 835-847.[27] K. van ’t Riet K, J.M. Smith, The behavior of gas-liquid mixtures near Rushton turbine blades, Chem. Eng.
Sci. 28 (1973) 1031-103.[28] T. Post, Understand the real world of mixing; Back to basics, Post mixing optimizations and solutions (2010)
(http://postmixing.com/publications/100315ceparticle.pdf; 08-04-2014)[29] E.M. Coen, R.G. Gilbert, Particle size distributions, in: J.M. Asua (Eds), Polymeric dispersions: Principle
and application, Kluwer Academic Publishers, The Netherlands, 1997, 67-78.[30] C.K.J. Keung, Emulsion polymerization of vinyl acetate: Particle size and molecular weight distributions,
(http://digitalcommons.mcmaster.ca/cgi/viewcontent.cgi?article=1059&context=opendissertations; 20-06-2012) .
[31] T. Kourti, Polymer latexes: production by homogeneous nucleation and methods for particle size determination, (http://digitalcommons.mcmaster.ca/cgi/viewcontent.cgi?article=3034&context=opendissertations; 20-06-2012) .
[32] N. Ohmura, K. Kitamoto, T. Yano, K. Kataoka, Novel operating method for controlling latex particle size distributions in emulsion polymerization of vinyl acetate, Ind. Eng. Chem. Res. 40 (2001) 5177-5183.
[33] A. Sood, Particle size distribution control in emulsion polymerization, J. Appl. Polym. Sci. 92 (2004) 2884-2902.
[34] N. Weber, M.M. Unterlass, K. Tauer, High ionic strength promotes the formation of spherical copolymer particles, Macromol. Chem. Phys. 212 (2011) 2071-2086.
[35] B.R. Morrison, R.G. Gilbert, Conditions for secondary particle formation in emulsion polymerization systems, Macromol. Symp. 92 (1995) 13-30.
[36] B.M. Budhlall, E.D. Sudol, V.L. Dimonie, A. Klein, M.S. El-Aasser, Role of grafting in the emulsion polymerization of vinyl acetate with poly(vinyl alcoholhol) as an emulsifier. I. Effect of the degree of blockiness on the kinetics and mechanism of grafting, J Polym Sci: Pt A Polym chem, 39 (20) (2001) 3633
[37] H. Lange, Emulsion polymerization of vinyl acetate with renewable raw materials as protective colloids, (http://kth.diva-portal.org/smash/record.jsf?pid=diva2:443050; 21-12-2012).
[38] W.J. Priest, Particle growth in the aqueous polymerization of vinyl acetate, J. Phys. Chem, 56 (1952) 1077-1082.
[39] A. Klein, C.H. Kuist, V.T. Stannet, Vinyl acetate emulsion polymerization. I. Effect of ionic strength and temperature on monomer solubility in the ionically stabilized polymer particle, J. Polym. Sci. Chem. Ed. 11 (1973) 2111-2126.
[40] J.W. Taylor, T.D. Klots, An applied approach to film formation the glass transition temperature evolution of latex films, 29th annual waterborne, high solids and powder coatings symposium in New Orleans (2002).
[41] D. Urban, K. Takamura, Polymer dispersions and their industrial applications, Wiley-VHC Verlagh GmbH & Co, Weinheim, 2002.
[42] D.M.C. Heymans, M.F. Daniel, Glass transition and film formation of veova/vinyl acetate latices; Role of water and co-solvents, Polymers for advanced technologies 6 (1995) 291.
[43] TAinstruments, Thermal analysis: An introduction (www.tainst.com; 21-12-2012).[44] J. Foreman, S.R. Sauerbrunn, C.L. Marcozzi, Exploring the sensitivity of thermal analysis techniques to the
glass transition, Application note TA-082 (www.tainst.com; 21-12-2012).[45] N.A. Bailey, J.N. Hay, D.M. Price, A study of enthalpic relaxation of poly(ethylene terephthalate) by
conventional and modulated temperature differential differential scanning calorimetry, Proceedings of the 27th conference of north american thermal analysis society in Savannah (1999).
[46] C. Bramsiepe, S. Sievers, T. Seifert, G.D. Stefanidis, D.G. Vlachos, H. Schnitzer, B. Muster, C. Brunner, J.P.M. Sanders, M.E. Bruins, G. Schembecker, Low-cost small scale processing technologies for production applications in various environments-mass produced factories, Chem. Eng. Process 51 (2012) 32-52.
[47] J.M. Ponce-Ortega, M.M. Al-Thubaiti, M.M. El-Halwagi, Process intensification: New understanding and systematic approach, Chem. Eng. Process 53 (2012) 63-75.
References[1] I. Skeist, Handbook of adhesives, 3rd edition, Chapman & Hall, New York, 1989[2] Freedoniagroup, World Emulsion Polymers: Industry Study with Forecasts for 2016 & 2021, 2012, (http://
www.freedoniagroup.com/brochure/29xx/2929smwe.pdf; 10-10-2012)[3] Chemical Market Resources, Inc. Worldwide vinyl acetate derivatives markets, technologies & trends 2006-
2011-2025 (http://cmrhoutex.com/media/mcs/MCS%20943-VAM%20Derivatives.pdf; 30-07-2013)[4] I. Piirma, J.L. Gardon, Emulsion polymerization, ACS symposium series: Volume 24, Washington, 1976[5] S. Carrà, A. Sliepcevich, A. Canevarolo, A. Carrà, Grafting and adsorption of poly(vinyl) alcohol in vinyl
acetate emulsion polymerization, Polymer, 46 (2005) 1379[6] F.K. Hansen, J. Ugelstad, Particle nucleation in emulsion polymerization: A theory for homogeneous
nucleation, J. Polym. Sci: Polym Chem, 16 (1978) 1953[7] J.P. Riggs, F. Rodriguez, Polymerization of acrylamide initiated by the persulfate-thiosulfate redox couple,
J. Polym. Sci. Chem Ed. 5 (1967) 3167-3181.[8] K.Pohl, F.Rodriguez, Adiabatic Polymerization of Acrylamide Using a Persulfate-Bisulfite Redox Couple, J
Appl Polym Sci, 1980, 26: 611[9] C. Liang, H.W. Su, Identification of sulfate and hydroxyl radicals in thermally activated persulfate, Ind. Eng.
Chem. Res. 48 (2009) 5558-5562.[10] FMC Corporation, Persulfates (http://www.fmcchemicals.com/TechDataSheetsMSDS/Persulfates.aspx;
20-06-2012).[11] E.P. Crematy, Endgroup studies of peroxodisulfate-initiated vinyl polymerization: Re-examinationof the
proposed mechanism of initiation in acid solution, J. Polym. Sci, Polym. Chem. Ed. 7 (1969) 3260.[12] I.M. Kolthof, I.K. Miller, The Chemistry of Persulfate. I. The Kinetics and Mechanism of the Decomposition
of the Persulfate Ion in Aqueous Medium, J Am Chem Soc, 73 (1951) 3055-3059[13] X. Ni, M.R. Mackley, A.P. Harvey, P. Stonestreet, M.H.I. Baird, N.V. Rama Rao NV, Mixing through oscillations
and pulsations: A guide to achieving process enhancements in the chemical and process industries, Trans IChemE, Part A, 81 (2003) 373-383
[14] X. Ni, Y. Zhang, I. Mustafa, An investigation of droplet size and size distribution in methylmethacrylate suspensions in a batch oscillatory-based reactor, Chemical Engineering Science, 53 (16) (1998) 2903-2919
[15] X. Ni, K.R. Murray, M. Zhang, D. Bennet, T. Howes, Polymer product engineering utilising oscillatory baffled reactors, Powder Technology 124 (2002) 281-286
[16] Schott, Borofloat 33 (http://www.schott.com/borofloat/english/attribute/thermic/index.html; 20-06-2012).[17] Atlas Steels, Grade datasheet: Stainless Steel 316 (http://www.atlassteels.com.au/site/pages/stainless-
steel-datasheets.php; 20-06-2012).[18] Euro Inox, Pickling and passivating of stainless steel, Material and applications series, Volume 24 (http://
www.euro-inox.org/pdf/map/Passivating_Pickling_EN.pdf; 20-06-2012)[19] Bionictechnology, Biocoating Metal 2 in 1 (http://www.bionictechnology.nl/gb-gb/products/1/Industrial/
biocoatmetaal2in1/Biocoating_Metal_2_in_1.html; 20-06-2012)[20] K. Kreilein, H. Hammer, Emulsion rapid polymerisation appts., useful esp. for vinyl chloride (1996)
DE4421949.[21] R. Rupaner, S. Lawrenz, G. Bauer, J. Dobbelaar, J. Nahstoll, F.J. Mueseler, A. Ferber, J. Hartmann, P.
Keller, M. Meister, J. Neutzner, G. Rehmer, Use of a single-stage or multistage stirrer to prepare polymers, (2001) US6252018.
[22] S. Kurita, K. Suzuki, M. Nakano, K. Maruo, Method of producing polymeric latex, WO/1993/022350A1[23] W. Baade, H.U. Moritz, K.H Reichert, Kinetics of high conversion polymerization of vinyl acetate. Effects of
mixing and reactor type on polymer properties, J. Appl. Polym. Sci. 27 (1982) 2249-2267[24] M.F. Kemmere, Batch emulsion polymerization: A chemical engineering approach, (http://alexandria.tue.nl/
extra3/proefschrift/boeken/9902484.pdf; 20-06-2012).[25] J. Meuldijk, M.F. Kemmere, S.V.W. de Lima, X.E.E. Reynhout, A.A.H. Drinkenburg, A.L. German, Some key
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
46 47
factors in emulsion polymerization process development, Polym. React. Eng. 11 (3) (2003) 259-276.[26] M. Nomura, M. Harada, W. Eguchi, S. Nagata, Effect of stirring on the emulsion polymerization of styrene,
J. Appl. Polym. Sci. 16 (1972) 835-847.[27] K. van ’t Riet K, J.M. Smith, The behavior of gas-liquid mixtures near Rushton turbine blades, Chem. Eng.
Sci. 28 (1973) 1031-103.[28] T. Post, Understand the real world of mixing; Back to basics, Post mixing optimizations and solutions (2010)
(http://postmixing.com/publications/100315ceparticle.pdf; 08-04-2014)[29] E.M. Coen, R.G. Gilbert, Particle size distributions, in: J.M. Asua (Eds), Polymeric dispersions: Principle
and application, Kluwer Academic Publishers, The Netherlands, 1997, 67-78.[30] C.K.J. Keung, Emulsion polymerization of vinyl acetate: Particle size and molecular weight distributions,
(http://digitalcommons.mcmaster.ca/cgi/viewcontent.cgi?article=1059&context=opendissertations; 20-06-2012) .
[31] T. Kourti, Polymer latexes: production by homogeneous nucleation and methods for particle size determination, (http://digitalcommons.mcmaster.ca/cgi/viewcontent.cgi?article=3034&context=opendissertations; 20-06-2012) .
[32] N. Ohmura, K. Kitamoto, T. Yano, K. Kataoka, Novel operating method for controlling latex particle size distributions in emulsion polymerization of vinyl acetate, Ind. Eng. Chem. Res. 40 (2001) 5177-5183.
[33] A. Sood, Particle size distribution control in emulsion polymerization, J. Appl. Polym. Sci. 92 (2004) 2884-2902.
[34] N. Weber, M.M. Unterlass, K. Tauer, High ionic strength promotes the formation of spherical copolymer particles, Macromol. Chem. Phys. 212 (2011) 2071-2086.
[35] B.R. Morrison, R.G. Gilbert, Conditions for secondary particle formation in emulsion polymerization systems, Macromol. Symp. 92 (1995) 13-30.
[36] B.M. Budhlall, E.D. Sudol, V.L. Dimonie, A. Klein, M.S. El-Aasser, Role of grafting in the emulsion polymerization of vinyl acetate with poly(vinyl alcoholhol) as an emulsifier. I. Effect of the degree of blockiness on the kinetics and mechanism of grafting, J Polym Sci: Pt A Polym chem, 39 (20) (2001) 3633
[37] H. Lange, Emulsion polymerization of vinyl acetate with renewable raw materials as protective colloids, (http://kth.diva-portal.org/smash/record.jsf?pid=diva2:443050; 21-12-2012).
[38] W.J. Priest, Particle growth in the aqueous polymerization of vinyl acetate, J. Phys. Chem, 56 (1952) 1077-1082.
[39] A. Klein, C.H. Kuist, V.T. Stannet, Vinyl acetate emulsion polymerization. I. Effect of ionic strength and temperature on monomer solubility in the ionically stabilized polymer particle, J. Polym. Sci. Chem. Ed. 11 (1973) 2111-2126.
[40] J.W. Taylor, T.D. Klots, An applied approach to film formation the glass transition temperature evolution of latex films, 29th annual waterborne, high solids and powder coatings symposium in New Orleans (2002).
[41] D. Urban, K. Takamura, Polymer dispersions and their industrial applications, Wiley-VHC Verlagh GmbH & Co, Weinheim, 2002.
[42] D.M.C. Heymans, M.F. Daniel, Glass transition and film formation of veova/vinyl acetate latices; Role of water and co-solvents, Polymers for advanced technologies 6 (1995) 291.
[43] TAinstruments, Thermal analysis: An introduction (www.tainst.com; 21-12-2012).[44] J. Foreman, S.R. Sauerbrunn, C.L. Marcozzi, Exploring the sensitivity of thermal analysis techniques to the
glass transition, Application note TA-082 (www.tainst.com; 21-12-2012).[45] N.A. Bailey, J.N. Hay, D.M. Price, A study of enthalpic relaxation of poly(ethylene terephthalate) by
conventional and modulated temperature differential differential scanning calorimetry, Proceedings of the 27th conference of north american thermal analysis society in Savannah (1999).
[46] C. Bramsiepe, S. Sievers, T. Seifert, G.D. Stefanidis, D.G. Vlachos, H. Schnitzer, B. Muster, C. Brunner, J.P.M. Sanders, M.E. Bruins, G. Schembecker, Low-cost small scale processing technologies for production applications in various environments-mass produced factories, Chem. Eng. Process 51 (2012) 32-52.
[47] J.M. Ponce-Ortega, M.M. Al-Thubaiti, M.M. El-Halwagi, Process intensification: New understanding and systematic approach, Chem. Eng. Process 53 (2012) 63-75.
References[1] I. Skeist, Handbook of adhesives, 3rd edition, Chapman & Hall, New York, 1989[2] Freedoniagroup, World Emulsion Polymers: Industry Study with Forecasts for 2016 & 2021, 2012, (http://
www.freedoniagroup.com/brochure/29xx/2929smwe.pdf; 10-10-2012)[3] Chemical Market Resources, Inc. Worldwide vinyl acetate derivatives markets, technologies & trends 2006-
2011-2025 (http://cmrhoutex.com/media/mcs/MCS%20943-VAM%20Derivatives.pdf; 30-07-2013)[4] I. Piirma, J.L. Gardon, Emulsion polymerization, ACS symposium series: Volume 24, Washington, 1976[5] S. Carrà, A. Sliepcevich, A. Canevarolo, A. Carrà, Grafting and adsorption of poly(vinyl) alcohol in vinyl
acetate emulsion polymerization, Polymer, 46 (2005) 1379[6] F.K. Hansen, J. Ugelstad, Particle nucleation in emulsion polymerization: A theory for homogeneous
nucleation, J. Polym. Sci: Polym Chem, 16 (1978) 1953[7] J.P. Riggs, F. Rodriguez, Polymerization of acrylamide initiated by the persulfate-thiosulfate redox couple,
J. Polym. Sci. Chem Ed. 5 (1967) 3167-3181.[8] K.Pohl, F.Rodriguez, Adiabatic Polymerization of Acrylamide Using a Persulfate-Bisulfite Redox Couple, J
Appl Polym Sci, 1980, 26: 611[9] C. Liang, H.W. Su, Identification of sulfate and hydroxyl radicals in thermally activated persulfate, Ind. Eng.
Chem. Res. 48 (2009) 5558-5562.[10] FMC Corporation, Persulfates (http://www.fmcchemicals.com/TechDataSheetsMSDS/Persulfates.aspx;
20-06-2012).[11] E.P. Crematy, Endgroup studies of peroxodisulfate-initiated vinyl polymerization: Re-examinationof the
proposed mechanism of initiation in acid solution, J. Polym. Sci, Polym. Chem. Ed. 7 (1969) 3260.[12] I.M. Kolthof, I.K. Miller, The Chemistry of Persulfate. I. The Kinetics and Mechanism of the Decomposition
of the Persulfate Ion in Aqueous Medium, J Am Chem Soc, 73 (1951) 3055-3059[13] X. Ni, M.R. Mackley, A.P. Harvey, P. Stonestreet, M.H.I. Baird, N.V. Rama Rao NV, Mixing through oscillations
and pulsations: A guide to achieving process enhancements in the chemical and process industries, Trans IChemE, Part A, 81 (2003) 373-383
[14] X. Ni, Y. Zhang, I. Mustafa, An investigation of droplet size and size distribution in methylmethacrylate suspensions in a batch oscillatory-based reactor, Chemical Engineering Science, 53 (16) (1998) 2903-2919
[15] X. Ni, K.R. Murray, M. Zhang, D. Bennet, T. Howes, Polymer product engineering utilising oscillatory baffled reactors, Powder Technology 124 (2002) 281-286
[16] Schott, Borofloat 33 (http://www.schott.com/borofloat/english/attribute/thermic/index.html; 20-06-2012).[17] Atlas Steels, Grade datasheet: Stainless Steel 316 (http://www.atlassteels.com.au/site/pages/stainless-
steel-datasheets.php; 20-06-2012).[18] Euro Inox, Pickling and passivating of stainless steel, Material and applications series, Volume 24 (http://
www.euro-inox.org/pdf/map/Passivating_Pickling_EN.pdf; 20-06-2012)[19] Bionictechnology, Biocoating Metal 2 in 1 (http://www.bionictechnology.nl/gb-gb/products/1/Industrial/
biocoatmetaal2in1/Biocoating_Metal_2_in_1.html; 20-06-2012)[20] K. Kreilein, H. Hammer, Emulsion rapid polymerisation appts., useful esp. for vinyl chloride (1996)
DE4421949.[21] R. Rupaner, S. Lawrenz, G. Bauer, J. Dobbelaar, J. Nahstoll, F.J. Mueseler, A. Ferber, J. Hartmann, P.
Keller, M. Meister, J. Neutzner, G. Rehmer, Use of a single-stage or multistage stirrer to prepare polymers, (2001) US6252018.
[22] S. Kurita, K. Suzuki, M. Nakano, K. Maruo, Method of producing polymeric latex, WO/1993/022350A1[23] W. Baade, H.U. Moritz, K.H Reichert, Kinetics of high conversion polymerization of vinyl acetate. Effects of
mixing and reactor type on polymer properties, J. Appl. Polym. Sci. 27 (1982) 2249-2267[24] M.F. Kemmere, Batch emulsion polymerization: A chemical engineering approach, (http://alexandria.tue.nl/
extra3/proefschrift/boeken/9902484.pdf; 20-06-2012).[25] J. Meuldijk, M.F. Kemmere, S.V.W. de Lima, X.E.E. Reynhout, A.A.H. Drinkenburg, A.L. German, Some key
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
48 49
AppendixTable A1: WPC (Sum 0-7 hours) and latex composition related responses.
Code Dry matter WPC Chemical composition
Calc.(mg/g)
Found(mg/g)
Recov.(%)
Sum0-7
hours(%)
pH Ethanal(mg/g)
VAM(mg/g)
Acetate (mmol)
Sulfate(mmol)
Thiosulfate(mmol)
LLL 522 516 98.8 16826 5.0 0.59 4.48 26.9 3.64 0.32LLH 523 518 99.1 34091 4.9 0.55 4.82 27.4 3.33 0.34LHL 521 512 98.3 66750 5.0 0.52 3.86 27.9 3.77 0.33LHH 520 517 99.3 40009 4.9 0.60 4.00 28.8 3.44 0.32HLL 522 512 98.1 33409 5.0 0.58 4.30 26.2 3.50 0.31HLH 520 518 99.6 39522 5.0 0.58 5.01 27.5 3.55 0.33HHL 524 516 98.5 63538 5.1 0.42 3.07 22.4 4.04 0.33HHH 525 516 98.3 48456 4.9 0.48 2.75 27.5 3.50 0.31CCC1 518 513 99.1 26367 5.0 0.25 1.85 25.6 3.26 0.33CCC2 521 516 99.0 36095 5.0 0.48 3.65 25.9 3.34 0.34CCC3 518 517 99.8 35573 5.1 0.47 3.56 25.7 3.25 0.34
Table A2: Viscosity and PSD related variables.Code Viscosity PSD
Peak Particles with diameter < 1mmMode Half width Amount Stability (Max – Min)
(mPa·s) (mm) (mm) (%) (%)
LLL 162 1.47 6.6 38 0.34LLH 148 1.33 5.9 42 0.28LHL 64 0.90 2.4 57 0.04LHH 86 0.88 3.4 60 0.09HLL 154 1.14 4.4 51 0.30HLH 192 1.10 4.1 54 0.19HHL 92 0.86 2.9 62 0.19HHH 102 0.85 3.1 67 0.42CCC1 124 0.77 4.8 65 0.26CCC2 124 0.96 5.5 57 0.11CCC3 106 0.83 5.0 60 0.13
[48] J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical engineering: Green engineering, Chemical Engineering Journal, 133 ( 2007) 7-30
[49] ACS Green Chemistry Institute, The twelve principles of green engineering,(www.acs.org; 18-03-2013)[50] ACS Green Chemistry Institute, The twelve principles of green chemistry (www.acs.org; 18-03-2013)[51] J.D. Browning , et al, Protein stabilized latex polymer emulsions, methods of making, and adhesives
containing such emulsions, (2010) US0099802[52] D.H. Craig, et al, Suspension polymerization of a vinyl monomer in the presence of (A) carboxymethyl
hydrophobically modified hydroxyethylcellulose (CMHMHEC) or (B) CMHMHEC in combination with an electrolyte or polyelectrolyte , (1989) US4868238
[53] P.F.T. Lambrechts, J.H.R, Van der Meeren, Method for preparing a water-containing vinyl acetate polymer dispersion, dispersion thus prepared and protective colloid used thereby (1982) US4322322
[54] H. Büsching, K. Friederich, H. Buxhover, E. Abrahams, R. Gossen, W. Schaper, , Adhesive dispersion for gumming in envelope machines, WO/1998/011171
[55] H. Buxhoffer, E. Abrahams-Meyer, R. Gossen, H. Büsching, K. Friederich, Rubber adhesive based on a stablized polyvinyl acetate dispersion, WO2000027943.
[56] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum adhesive based on a filled polymer dispersion, WO2006094594A1.
[57] K.L. Hsieh, L.I. Tong, H.P. Chiu, H.Y. Yeh, Optimization of a multi-response problem in Taguchi’s dynamic system, Comput. Ind. Eng. 49 (2005) 556–571.
[58] F. Luo, H. Sun, T. Geng, N. Qi, Application of Taguchi’s method in the optimization of bridging efficiency between confluent and fresh microcarriers in bead-to-bead transfer of vero cells, Biotechnol. Lett., 30 (2008), 645–649.
[59] Experimental design and Taguchi: (http://homepage.ntlworld.com/s.orszulik/index.html; 20-06-2012).[60] H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/
bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).[61] W.D. Hergeth, W. Lebek, R. Kakuschke, K Schmutzler, Particle formation in emulsion polymerization, 1
Oligomers in emulsion polymerization of vinyl acetate, Makromol. Chem,, 192 (1991) 2265-2275.[62] D.A. House,. Kinetics and mechanism of oxidations by peroxydisulfate, Chem. Rev, 62 (1962) 185[63] C.Liang, C.J. Bruell, M.C. Marley, K.L. Sperry, Persulfate oxidation for in situ remediation of TCE. I. Activated
by ferrous ion with and without a persulfate-thiosulfate redox couple, Chemosphere 55 (2004) 1213-1223[64] S.H. Samaha, H.E. Nasr, A. Hebeish, Synthesis and Characterization of Starch-Poly(vinyl Acetate) Graft
Copolymers and their Saponified Form, Journal of polymer Research,12 (2005) 343.[65] D. Braun, H. Cherdron, H. Ritter, Polymer synthesis: Theory and Practice: Fundamentals, Methods,
Experiments, Third edition, Springer-Verlag, Berlin, 2001.[66] K.B. Renuka Devi, R. Madivanane, Normal coordinate analysis of polyvinyl acetate, Engineering science
and technology: An international journal, 2 (4) (2012) 795-799[67] K.M. Munmaya, Y. Yusuf, Handbook of Vinyl Polymers Radical Polymerization, Process and Technology,
2nd Ed., CRC press, Boca Raton, 2009.[68] D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization
of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerizations aspects
48 49
AppendixTable A1: WPC (Sum 0-7 hours) and latex composition related responses.
Code Dry matter WPC Chemical composition
Calc.(mg/g)
Found(mg/g)
Recov.(%)
Sum0-7
hours(%)
pH Ethanal(mg/g)
VAM(mg/g)
Acetate (mmol)
Sulfate(mmol)
Thiosulfate(mmol)
LLL 522 516 98.8 16826 5.0 0.59 4.48 26.9 3.64 0.32LLH 523 518 99.1 34091 4.9 0.55 4.82 27.4 3.33 0.34LHL 521 512 98.3 66750 5.0 0.52 3.86 27.9 3.77 0.33LHH 520 517 99.3 40009 4.9 0.60 4.00 28.8 3.44 0.32HLL 522 512 98.1 33409 5.0 0.58 4.30 26.2 3.50 0.31HLH 520 518 99.6 39522 5.0 0.58 5.01 27.5 3.55 0.33HHL 524 516 98.5 63538 5.1 0.42 3.07 22.4 4.04 0.33HHH 525 516 98.3 48456 4.9 0.48 2.75 27.5 3.50 0.31CCC1 518 513 99.1 26367 5.0 0.25 1.85 25.6 3.26 0.33CCC2 521 516 99.0 36095 5.0 0.48 3.65 25.9 3.34 0.34CCC3 518 517 99.8 35573 5.1 0.47 3.56 25.7 3.25 0.34
Table A2: Viscosity and PSD related variables.Code Viscosity PSD
Peak Particles with diameter < 1mmMode Half width Amount Stability (Max – Min)
(mPa·s) (mm) (mm) (%) (%)
LLL 162 1.47 6.6 38 0.34LLH 148 1.33 5.9 42 0.28LHL 64 0.90 2.4 57 0.04LHH 86 0.88 3.4 60 0.09HLL 154 1.14 4.4 51 0.30HLH 192 1.10 4.1 54 0.19HHL 92 0.86 2.9 62 0.19HHH 102 0.85 3.1 67 0.42CCC1 124 0.77 4.8 65 0.26CCC2 124 0.96 5.5 57 0.11CCC3 106 0.83 5.0 60 0.13
[48] J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical engineering: Green engineering, Chemical Engineering Journal, 133 ( 2007) 7-30
[49] ACS Green Chemistry Institute, The twelve principles of green engineering,(www.acs.org; 18-03-2013)[50] ACS Green Chemistry Institute, The twelve principles of green chemistry (www.acs.org; 18-03-2013)[51] J.D. Browning , et al, Protein stabilized latex polymer emulsions, methods of making, and adhesives
containing such emulsions, (2010) US0099802[52] D.H. Craig, et al, Suspension polymerization of a vinyl monomer in the presence of (A) carboxymethyl
hydrophobically modified hydroxyethylcellulose (CMHMHEC) or (B) CMHMHEC in combination with an electrolyte or polyelectrolyte , (1989) US4868238
[53] P.F.T. Lambrechts, J.H.R, Van der Meeren, Method for preparing a water-containing vinyl acetate polymer dispersion, dispersion thus prepared and protective colloid used thereby (1982) US4322322
[54] H. Büsching, K. Friederich, H. Buxhover, E. Abrahams, R. Gossen, W. Schaper, , Adhesive dispersion for gumming in envelope machines, WO/1998/011171
[55] H. Buxhoffer, E. Abrahams-Meyer, R. Gossen, H. Büsching, K. Friederich, Rubber adhesive based on a stablized polyvinyl acetate dispersion, WO2000027943.
[56] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum adhesive based on a filled polymer dispersion, WO2006094594A1.
[57] K.L. Hsieh, L.I. Tong, H.P. Chiu, H.Y. Yeh, Optimization of a multi-response problem in Taguchi’s dynamic system, Comput. Ind. Eng. 49 (2005) 556–571.
[58] F. Luo, H. Sun, T. Geng, N. Qi, Application of Taguchi’s method in the optimization of bridging efficiency between confluent and fresh microcarriers in bead-to-bead transfer of vero cells, Biotechnol. Lett., 30 (2008), 645–649.
[59] Experimental design and Taguchi: (http://homepage.ntlworld.com/s.orszulik/index.html; 20-06-2012).[60] H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/
bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).[61] W.D. Hergeth, W. Lebek, R. Kakuschke, K Schmutzler, Particle formation in emulsion polymerization, 1
Oligomers in emulsion polymerization of vinyl acetate, Makromol. Chem,, 192 (1991) 2265-2275.[62] D.A. House,. Kinetics and mechanism of oxidations by peroxydisulfate, Chem. Rev, 62 (1962) 185[63] C.Liang, C.J. Bruell, M.C. Marley, K.L. Sperry, Persulfate oxidation for in situ remediation of TCE. I. Activated
by ferrous ion with and without a persulfate-thiosulfate redox couple, Chemosphere 55 (2004) 1213-1223[64] S.H. Samaha, H.E. Nasr, A. Hebeish, Synthesis and Characterization of Starch-Poly(vinyl Acetate) Graft
Copolymers and their Saponified Form, Journal of polymer Research,12 (2005) 343.[65] D. Braun, H. Cherdron, H. Ritter, Polymer synthesis: Theory and Practice: Fundamentals, Methods,
Experiments, Third edition, Springer-Verlag, Berlin, 2001.[66] K.B. Renuka Devi, R. Madivanane, Normal coordinate analysis of polyvinyl acetate, Engineering science
and technology: An international journal, 2 (4) (2012) 795-799[67] K.M. Munmaya, Y. Yusuf, Handbook of Vinyl Polymers Radical Polymerization, Process and Technology,
2nd Ed., CRC press, Boca Raton, 2009.[68] D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization
of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837
50
CHAPTER 3A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
Table A3: Amount of homopolymer, grafted polymer and f for all prepared latexes.Code Cold extraction Hot extraction pVAc composition
InsolublepVAc
(mg/g)
SolublepVAc
(mg/g)
Precipitate cold extraction
(%)
pVAc incold
precipitate(mg/g)
Total(mg/g)
Homo-polymer(mg/g)
Graftedpolymer(mg/g)
f(fraction)
LLL 196 320 35.2 69 468 389 79 0.17LLH 115 403 53.0 61 469 464 5 0.01LHL 314 198 54.3 170 466 368 98 0.21LHH 142 374 55.8 80 466 454 12 0.03HLL 177 335 31.6 56 468 391 77 0.16HLH 84 434 37.3 31 466 465 1 0.00HHL 338 178 29.9 101 470 279 191 0.41HHH 100 416 45.2 45 471 461 9 0.02CCC1 234 280 39.2 92 464 371 92 0.20CCC2 246 269 43.1 106 466 376 91 0.19CCC3 253 263 45.1 114 463 378 85 0.18
Table A4: Thermal transitions based on total heat flow and reversing heat flow.Code Thermal transitions
Total heat flow (mDSC) Reversed heat flow (mDSC)Tg,onset
(°C)Tg,inflection
(°C)Tg,endset
(°C)DTg
(°C)Tg,onset
(°C)Tg,inflection
(°C)Tg,endset
(°C)DTg
(°C)
LLL 5.8 8.2 9.7 3.9 9.5 12.0 14.6 5.2LLH 6.2 7.8 9.2 3.0 9.8 11.4 14.4 4.6LHL 6.4 10.3 11.1 4.7 11.1 14.2 16.0 4.9LHH 6.4 8.2 9.1 2.7 9.5 11.9 13.8 4.3HLL 6.5 8.6 10.1 3.5 10.7 12.5 15.5 4.8HLH 6.8 8.5 9.4 2.6 10.3 12.3 14.4 4.1HHL 8.4 10.5 11.6 3.2 11.8 14.4 16.3 4.4HHH 6.8 9.3 9.7 2.9 10.0 11.7 14.8 4.7CCC1 6.6 9.5 10.4 3.8 11.2 12.8 15.7 4.5CCC2 8.3 10.2 10.5 2.2 11.7 13.5 15.6 3.9CCC3 7.0 9.9 10.4 3.4 10.9 13.0 15.7 4.8
50
CHAPTER 3A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
Table A3: Amount of homopolymer, grafted polymer and f for all prepared latexes.Code Cold extraction Hot extraction pVAc composition
InsolublepVAc
(mg/g)
SolublepVAc
(mg/g)
Precipitate cold extraction
(%)
pVAc incold
precipitate(mg/g)
Total(mg/g)
Homo-polymer(mg/g)
Graftedpolymer(mg/g)
f(fraction)
LLL 196 320 35.2 69 468 389 79 0.17LLH 115 403 53.0 61 469 464 5 0.01LHL 314 198 54.3 170 466 368 98 0.21LHH 142 374 55.8 80 466 454 12 0.03HLL 177 335 31.6 56 468 391 77 0.16HLH 84 434 37.3 31 466 465 1 0.00HHL 338 178 29.9 101 470 279 191 0.41HHH 100 416 45.2 45 471 461 9 0.02CCC1 234 280 39.2 92 464 371 92 0.20CCC2 246 269 43.1 106 466 376 91 0.19CCC3 253 263 45.1 114 463 378 85 0.18
Table A4: Thermal transitions based on total heat flow and reversing heat flow.Code Thermal transitions
Total heat flow (mDSC) Reversed heat flow (mDSC)Tg,onset
(°C)Tg,inflection
(°C)Tg,endset
(°C)DTg
(°C)Tg,onset
(°C)Tg,inflection
(°C)Tg,endset
(°C)DTg
(°C)
LLL 5.8 8.2 9.7 3.9 9.5 12.0 14.6 5.2LLH 6.2 7.8 9.2 3.0 9.8 11.4 14.4 4.6LHL 6.4 10.3 11.1 4.7 11.1 14.2 16.0 4.9LHH 6.4 8.2 9.1 2.7 9.5 11.9 13.8 4.3HLL 6.5 8.6 10.1 3.5 10.7 12.5 15.5 4.8HLH 6.8 8.5 9.4 2.6 10.3 12.3 14.4 4.1HHL 8.4 10.5 11.6 3.2 11.8 14.4 16.3 4.4HHH 6.8 9.3 9.7 2.9 10.0 11.7 14.8 4.7CCC1 6.6 9.5 10.4 3.8 11.2 12.8 15.7 4.5CCC2 8.3 10.2 10.5 2.2 11.7 13.5 15.6 3.9CCC3 7.0 9.9 10.4 3.4 10.9 13.0 15.7 4.8
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
52 53
Figure 1: The two proposed patterns for the conversion of glucose into HMF by dehydration [16].
Figure 2: Proposed reaction mechanism for the conversion of HMF into humins and levulinic acid [17].
Starch becomes yellow when exposed to pyrodextrination conditions and this colouration is related to a change in the amount of conjugated double bonds. The latter is associated with radical scavenging properties and could therefore have a considerable impact on the efficiency of a free radical polymerization process [23]. Therefore, pyrodextrins with a lower degree of yellowness are preferred if there is a choice between pyrodextrins with similar intrinsic viscosities and different levels of yellowness (Figure 3) [20]. The intrinsic viscosity [η] is a measure of a solute’s contribution to the viscosity of a solution (Equation 1) [24]:
Equation1:
where c denotes the concentration of the dissolved polymer and ηsp stands for specific viscosity of the solution. The latter can be calculated from the actual viscosity of the solution (η) and the viscosity of the pure solvent (η0) (Equation 2) [24]:
Equation 2:
-
-
- - -
-
HMF
Humins (polymer)
Levulinic acid+2
-
-
lim sp
cc 0
AbstractThe influence of the level of pyrodextrination of potato starch on the performance as a protective colloid for polyvinyl acetate latexes was investigated systematically. A simplified Taguchi design of experiments was used to investigate the influence of several different process variables of the pyrodextrins preparation process (namely amount of added hydrochloric acid (HCl = 15, 18 and 21 mmol/kg starch), level of pre-drying (PDM = 15, 30 and 45 minutes at 90°C), dextrination time (DTH = 3, 5 and 7 hours at 160°C)) on the final properties of the pyrodextrins. The pyrodextrination process appears to be controlled by the factors HCl and DTH mainly. The variation in reaction conditions resulted in pyrodextrins with distinct differences in colouration and viscosity level. The protective colloid properties of the pyrodextrins were evaluated in vinyl acetate free-radical polymerization. All latexes were generated with the same polymerization procedure and showed similar monomer conversion levels and dry matter recoveries above 97 wt %. The process variables used during pyrodextrination only had a limited influence on the glass transition temperature of the latexes prepared. However, other process and product composition related variables (e.g. power consumption, conversion, recovery and grafting) show pronounced differences that can be related to the pyrodextrination process factors HCl and DTH mainly. Latexes with a monodisperse particle size distribution require a higher energy input during processing than their polydisperse counterparts.
IntroductionThe commercial application of (pyro)dextrins as protective colloids in free radical polymerizations dates back to at least five decades. However, it still constitutes a relevant research topic, at least at industrial level, as testified by the number of publications in the recent past [1-9]. Polyvinyl acetate (pVAc) latexes stabilized with pyrodextrins are suitable for formulating (re-moistenable) adhesives characterized by dry matter contents up to 70 wt%. Pyrodextrins can be applied in two different ways in emulsion polymerization, i.e. either by adding the total amount before or partly after the polymerization reaction. Latexes with similar properties can be produced in both ways and also with different types of pyrodextrins. The principle of pyrodextrination involves a pre-defined level of acidification of a starch source, moisture removal and exposure to heat [10]. Acidic hydrolysis and transglycosylation are believed to be the most important mechanisms for changing the structure of starch during pyrodextrination, both resulting in relatively small starch fragments with a high level of branching [11-15]. The colouration of starch during a pyrodextrination process is associated with the formation of conjugated double bonds, which can be generated in carbohydrates by dehydration of the carbohydrate moieties and aldol additions/condensations. The amount of reducing sugars increases with the extent of starch hydrolysis, with the subsequent dehydration resulting in hydroxyl methyl furfural (HMF) or related structures (Figure 1). HMF can be converted into other compounds at hydrothermal conditions (Figure 2). The intermediate 2,5-dioxo-6-hydroxy-hexanal contains α-hydrogens, next to a carbonyl group, and is believed to form complex polymers by aldol addition/condensation. The polymers formed are called humins and are characterized by a relatively high amount of conjugated double bonds [16-18]. Caramelization can occur at pyrodextrination conditions as well and this reaction also results in starch with conjugated double bonds [19-22].
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
52 53
Figure 1: The two proposed patterns for the conversion of glucose into HMF by dehydration [16].
Figure 2: Proposed reaction mechanism for the conversion of HMF into humins and levulinic acid [17].
Starch becomes yellow when exposed to pyrodextrination conditions and this colouration is related to a change in the amount of conjugated double bonds. The latter is associated with radical scavenging properties and could therefore have a considerable impact on the efficiency of a free radical polymerization process [23]. Therefore, pyrodextrins with a lower degree of yellowness are preferred if there is a choice between pyrodextrins with similar intrinsic viscosities and different levels of yellowness (Figure 3) [20]. The intrinsic viscosity [η] is a measure of a solute’s contribution to the viscosity of a solution (Equation 1) [24]:
Equation1:
where c denotes the concentration of the dissolved polymer and ηsp stands for specific viscosity of the solution. The latter can be calculated from the actual viscosity of the solution (η) and the viscosity of the pure solvent (η0) (Equation 2) [24]:
Equation 2:
-
-
- - -
-
HMF
Humins (polymer)
Levulinic acid+2
-
-
lim sp
cc 0
AbstractThe influence of the level of pyrodextrination of potato starch on the performance as a protective colloid for polyvinyl acetate latexes was investigated systematically. A simplified Taguchi design of experiments was used to investigate the influence of several different process variables of the pyrodextrins preparation process (namely amount of added hydrochloric acid (HCl = 15, 18 and 21 mmol/kg starch), level of pre-drying (PDM = 15, 30 and 45 minutes at 90°C), dextrination time (DTH = 3, 5 and 7 hours at 160°C)) on the final properties of the pyrodextrins. The pyrodextrination process appears to be controlled by the factors HCl and DTH mainly. The variation in reaction conditions resulted in pyrodextrins with distinct differences in colouration and viscosity level. The protective colloid properties of the pyrodextrins were evaluated in vinyl acetate free-radical polymerization. All latexes were generated with the same polymerization procedure and showed similar monomer conversion levels and dry matter recoveries above 97 wt %. The process variables used during pyrodextrination only had a limited influence on the glass transition temperature of the latexes prepared. However, other process and product composition related variables (e.g. power consumption, conversion, recovery and grafting) show pronounced differences that can be related to the pyrodextrination process factors HCl and DTH mainly. Latexes with a monodisperse particle size distribution require a higher energy input during processing than their polydisperse counterparts.
IntroductionThe commercial application of (pyro)dextrins as protective colloids in free radical polymerizations dates back to at least five decades. However, it still constitutes a relevant research topic, at least at industrial level, as testified by the number of publications in the recent past [1-9]. Polyvinyl acetate (pVAc) latexes stabilized with pyrodextrins are suitable for formulating (re-moistenable) adhesives characterized by dry matter contents up to 70 wt%. Pyrodextrins can be applied in two different ways in emulsion polymerization, i.e. either by adding the total amount before or partly after the polymerization reaction. Latexes with similar properties can be produced in both ways and also with different types of pyrodextrins. The principle of pyrodextrination involves a pre-defined level of acidification of a starch source, moisture removal and exposure to heat [10]. Acidic hydrolysis and transglycosylation are believed to be the most important mechanisms for changing the structure of starch during pyrodextrination, both resulting in relatively small starch fragments with a high level of branching [11-15]. The colouration of starch during a pyrodextrination process is associated with the formation of conjugated double bonds, which can be generated in carbohydrates by dehydration of the carbohydrate moieties and aldol additions/condensations. The amount of reducing sugars increases with the extent of starch hydrolysis, with the subsequent dehydration resulting in hydroxyl methyl furfural (HMF) or related structures (Figure 1). HMF can be converted into other compounds at hydrothermal conditions (Figure 2). The intermediate 2,5-dioxo-6-hydroxy-hexanal contains α-hydrogens, next to a carbonyl group, and is believed to form complex polymers by aldol addition/condensation. The polymers formed are called humins and are characterized by a relatively high amount of conjugated double bonds [16-18]. Caramelization can occur at pyrodextrination conditions as well and this reaction also results in starch with conjugated double bonds [19-22].
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
54 55
In this study, a typical pyrodextrination process for preparing a product with protective colloid characteristics was selected and the process parameters pre-drying (PDM), hydrochloric acid (HCl) and dextrination time (DTH) were varied to generate pyrodextrins with different structures and properties. The obtained pyrodextrins were used as a protective colloid in a vinyl acetate monomer (VAM) based polymerization. Sodium persulfate (SPS) was mainly used in thermal dissociation mode and the pH was kept in the range of 5 – 7 during polymerization with the help of sodium bicarbonate (SBC). Remnant SPS was converted by sodium thiosulfate (STS) after the polymerization was finished. The impact of the applied changes on process and product characteristics was evaluated. The general strategy is depicted in Scheme 1. The overall process consists of two steps, i.e. pyrodextrination and polymerization processes. These can be in principle statistically analyzed in an independent manner, but it is also possible to link the pyrodextrin variables HCl, PDM and DTH directly to the process and product characteristics of the latex. The latter does not require extensive characterization of the pyrodextrins used, and this is desirable due to their very complex chemical nature, and only demands a fixed latex preparation procedure. This integrative approach, where the input variables of the pyrodextrination are statistically related to the process and product characteristics, besides its novelty, is preferred and also used in the present work.
Scheme 1: General depiction of the process steps, and most important variables, involved for making a pyrodextrin stabilized pVAc latex (WPC = Water bath power consumption and PSD = particle size distribution).
ExperimentalMaterialsThe potato starch used in this study is commercially available under the trade name Potato starch Food grade (AVEBE U.A). The vinyl acetate monomer (VAM) was purchased from ACROS and contains 3-30 ppm hydroquinone. Analytical reagent grade sodium persulfate (SPS) was supplied by VWR International. Hydrochloric acid (HCl), Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) were of pro analyse quality and were purchased from Merck. All ingredients were used without additional purification and the solvent was demineralised water in all cases.
Pyrodextrination of starch1 kg potato starch (20 wt% moisture) was placed in the bowl of a Hobart mixer. A given amount of HCl (L = 15 mmol; C = 18 mmol and H = 21 mmol) was dissolved in 160 ml demi-water and then slowly added to the starch during mixing. The obtained mixture was stored for 16 hours at 8°C in a closed plastic bag. The mixture was then pre-dryed twice in a Retch
Product characteristics(e.g. PSD)
VAM
Starch Product characteristics Process characterisics(e.g. YI-E313) (e.g. WPC)
Pyrodextrin PolymerizationPyrodextrination Latex
PDMHCl DTH SPC/SBCWater
Figure 3: Yellow index 400nm (YI-400 nm) and the weight average of [η] as function of derivatization time [20].
Pyrodextrins used as protective colloids are usually intense yellow and can be quantified by the yellow Index E313 (YI-E313) [25]. However, the Commission Internationale de l’Eclairage (CIE) defined the L*a*b* colour space and this is a more discriminative approach for determining differences in colouration between pyrodextrins. This method relies on colour splitting into the components L* (black - white shift or intensity), a* (green - red shift) and b* (blue - yellow shift) respectively [26]. The pyrodextrination also has a considerable impact on the level of viscosity of the product after dissolution. Changes in viscosity, of either pyrodextrins in water or the final latex, can be measured by absolute (e.g. dynamic rotational; Brookfield) or kinematic (e.g. flow through an orifice; Ostwald) based viscometers and the latter is typically ten times more precise then the former [27]. Differences in amount of hydrocholoric acid (HCl) present in the pyrodextrin might be of influence on the polymerization reaction as well. Moreover, changes in pH and ionic strength can influence the radical and particle formation process and chloride is reported as a radical scavenger [28-29].The demand of more sustainable products is steadily increasing and the general rules of green chemistry and engineering are useful guidelines to improve current production processes and products [30-32]. Pyrodextrination and free radical polymerizations are both flexible production processes with respect to changes in reaction conditions (e.g. processing temperature and reagent concentrations), but characterized by high processing temperatures for common free radical polymerizations (i.e. 50 to 80 °C) and even higher processing temperatures for pyrodextrination (i.e 75 to 250 °C) [2,10]. Pyrodextrins can be prepared by low reaction temperatures, but the corresponding products are believed to be less suitable as protective colloid in free radical polymerizations. The level of cold water solubility of the pyrodextrin is usually proportional to the total heat exposure and a change of this variable might introduce problems during polymerization and/or application if no precautions are taken [10]. Systematic studies on the impact of the processing conditions on the pyrodextrins performance as protective colloids is lacking in the open literature despite the fact that it is an absolute necessity for proper optimization of pyrodextrin stabilized pVAc latexes.
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
54 55
In this study, a typical pyrodextrination process for preparing a product with protective colloid characteristics was selected and the process parameters pre-drying (PDM), hydrochloric acid (HCl) and dextrination time (DTH) were varied to generate pyrodextrins with different structures and properties. The obtained pyrodextrins were used as a protective colloid in a vinyl acetate monomer (VAM) based polymerization. Sodium persulfate (SPS) was mainly used in thermal dissociation mode and the pH was kept in the range of 5 – 7 during polymerization with the help of sodium bicarbonate (SBC). Remnant SPS was converted by sodium thiosulfate (STS) after the polymerization was finished. The impact of the applied changes on process and product characteristics was evaluated. The general strategy is depicted in Scheme 1. The overall process consists of two steps, i.e. pyrodextrination and polymerization processes. These can be in principle statistically analyzed in an independent manner, but it is also possible to link the pyrodextrin variables HCl, PDM and DTH directly to the process and product characteristics of the latex. The latter does not require extensive characterization of the pyrodextrins used, and this is desirable due to their very complex chemical nature, and only demands a fixed latex preparation procedure. This integrative approach, where the input variables of the pyrodextrination are statistically related to the process and product characteristics, besides its novelty, is preferred and also used in the present work.
Scheme 1: General depiction of the process steps, and most important variables, involved for making a pyrodextrin stabilized pVAc latex (WPC = Water bath power consumption and PSD = particle size distribution).
ExperimentalMaterialsThe potato starch used in this study is commercially available under the trade name Potato starch Food grade (AVEBE U.A). The vinyl acetate monomer (VAM) was purchased from ACROS and contains 3-30 ppm hydroquinone. Analytical reagent grade sodium persulfate (SPS) was supplied by VWR International. Hydrochloric acid (HCl), Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) were of pro analyse quality and were purchased from Merck. All ingredients were used without additional purification and the solvent was demineralised water in all cases.
Pyrodextrination of starch1 kg potato starch (20 wt% moisture) was placed in the bowl of a Hobart mixer. A given amount of HCl (L = 15 mmol; C = 18 mmol and H = 21 mmol) was dissolved in 160 ml demi-water and then slowly added to the starch during mixing. The obtained mixture was stored for 16 hours at 8°C in a closed plastic bag. The mixture was then pre-dryed twice in a Retch
Product characteristics(e.g. PSD)
VAM
Starch Product characteristics Process characterisics(e.g. YI-E313) (e.g. WPC)
Pyrodextrin PolymerizationPyrodextrination Latex
PDMHCl DTH SPC/SBCWater
Figure 3: Yellow index 400nm (YI-400 nm) and the weight average of [η] as function of derivatization time [20].
Pyrodextrins used as protective colloids are usually intense yellow and can be quantified by the yellow Index E313 (YI-E313) [25]. However, the Commission Internationale de l’Eclairage (CIE) defined the L*a*b* colour space and this is a more discriminative approach for determining differences in colouration between pyrodextrins. This method relies on colour splitting into the components L* (black - white shift or intensity), a* (green - red shift) and b* (blue - yellow shift) respectively [26]. The pyrodextrination also has a considerable impact on the level of viscosity of the product after dissolution. Changes in viscosity, of either pyrodextrins in water or the final latex, can be measured by absolute (e.g. dynamic rotational; Brookfield) or kinematic (e.g. flow through an orifice; Ostwald) based viscometers and the latter is typically ten times more precise then the former [27]. Differences in amount of hydrocholoric acid (HCl) present in the pyrodextrin might be of influence on the polymerization reaction as well. Moreover, changes in pH and ionic strength can influence the radical and particle formation process and chloride is reported as a radical scavenger [28-29].The demand of more sustainable products is steadily increasing and the general rules of green chemistry and engineering are useful guidelines to improve current production processes and products [30-32]. Pyrodextrination and free radical polymerizations are both flexible production processes with respect to changes in reaction conditions (e.g. processing temperature and reagent concentrations), but characterized by high processing temperatures for common free radical polymerizations (i.e. 50 to 80 °C) and even higher processing temperatures for pyrodextrination (i.e 75 to 250 °C) [2,10]. Pyrodextrins can be prepared by low reaction temperatures, but the corresponding products are believed to be less suitable as protective colloid in free radical polymerizations. The level of cold water solubility of the pyrodextrin is usually proportional to the total heat exposure and a change of this variable might introduce problems during polymerization and/or application if no precautions are taken [10]. Systematic studies on the impact of the processing conditions on the pyrodextrins performance as protective colloids is lacking in the open literature despite the fact that it is an absolute necessity for proper optimization of pyrodextrin stabilized pVAc latexes.
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
56 57
Figure 4: Schematic representation of the polymerization reactor used.
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM dosage was also monitored with a balance. One syringe pump was used to add a premix of SPS and SBC in water and another syringe pump was used to add STS in water after the actual polymerization was finished.
Polymerization procedureThe polymerization reactor was filled with 257.5-258.1 g of demineralized water and 33.8-34.4 g of a potato starch based pyrodextrin (moisture content: 1.1-2.9 wt %) was added to reach a pyrodextrin concentration of 10 wt % (with respect to the water amount) in the final latex. Automatic mixing (0-0.75 hours: 240 RPM; 0.75-9.0 hours: 120 RPM) was started after one minute of gentle manual homogenization. The applied temperature profile and dosage protocols of VAM, SPS/SBC and STS are given in Figure 5. VAM was used without inhibitor removal in order to closely resemble experimental conditions applicable at industrial scale. A total of 200 g VAM was added in all cases with a pre-dosage of 15 g VAM. The actual dosage was monitored in time and the total amount of VAM added was used for mass balance calculations. 1.5 g SPS and 2.0 g SBC were dissolved together in 46.5 g demineralised water and 30.5 ml of this mixture was added during the polymerization. The actual addition started after 104 minutes with a pre-dosage of 5 ml (5 ml/minutes) followed by 25.5 ml with a dosage speed of 5.8 ml/hour. The reducing agent was 0.3 mM STS and 3 ml was added with 3 ml/hour after 7.5 hours after the preparation procedure was started.
fluid bed dryer for 60 minutes at 60°C and for several minutes at 90 °C (L = 15 minutes; C = 30 minutes and H = 45 minutes) and immediately added to a rotating reactor with a constant temperature of 160°C. The product was removed from the reactor at x hours after addition (L = 3 hours; C = 5 hours and H = 7 hours) and stored at room temperature in a plastic bag after cooling down on a piece of paper for 2 hours.
Experimental designA Taguchi L8 orthogonal design allows the evaluation of the three factors (i.e. HCl, PDM and DTH) and their interactions [9]. The effect of a factor, or an interaction, is in the simplified Taguchi method calculated by multiplying the measured value with the corresponding number in the Taguchi L8 matrix (Table 1). The effect of a factor, or an interaction, is defined as the sum of the obtained values of the eight runs performed after this multiplication. The Taguchi L8 design was augmented with a centre point (C). This setting was executed in triplicate in order to get an impression of the overall level of noise in the pyrodextrination procedure [9,33].
Table 1: Taguchi L8 matrix for three factors and their interactions.Run Code Factors or interactions
AT BT ATBT CT ATCT BTCT ATBTCT
1 LLL -1 -1 -1 -1 -1 -1 -12 LLH -1 -1 -1 1 1 1 13 LHL -1 1 1 -1 -1 1 14 LHH -1 1 1 1 1 -1 -15 HLL 1 -1 1 -1 1 -1 16 HLH 1 -1 1 1 -1 1 -17 HHL 1 1 -1 -1 1 1 -18 HHH 1 1 -1 1 -1 -1 1
Polymerization equipmentA double-jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles) with a width of 1 cm and an outer dimension of 10.5x7 cm (height x diameter)) was used. A lid made of borosilicate glass with several connection points was placed on top and the reactor was completely insulated with radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the temperature of the headspace in the reactor (HST). The feeding lines of VAM and the SPS/SBC mixture were placed outside the reflux region with the aid of an accessory to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 4).
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
56 57
Figure 4: Schematic representation of the polymerization reactor used.
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM dosage was also monitored with a balance. One syringe pump was used to add a premix of SPS and SBC in water and another syringe pump was used to add STS in water after the actual polymerization was finished.
Polymerization procedureThe polymerization reactor was filled with 257.5-258.1 g of demineralized water and 33.8-34.4 g of a potato starch based pyrodextrin (moisture content: 1.1-2.9 wt %) was added to reach a pyrodextrin concentration of 10 wt % (with respect to the water amount) in the final latex. Automatic mixing (0-0.75 hours: 240 RPM; 0.75-9.0 hours: 120 RPM) was started after one minute of gentle manual homogenization. The applied temperature profile and dosage protocols of VAM, SPS/SBC and STS are given in Figure 5. VAM was used without inhibitor removal in order to closely resemble experimental conditions applicable at industrial scale. A total of 200 g VAM was added in all cases with a pre-dosage of 15 g VAM. The actual dosage was monitored in time and the total amount of VAM added was used for mass balance calculations. 1.5 g SPS and 2.0 g SBC were dissolved together in 46.5 g demineralised water and 30.5 ml of this mixture was added during the polymerization. The actual addition started after 104 minutes with a pre-dosage of 5 ml (5 ml/minutes) followed by 25.5 ml with a dosage speed of 5.8 ml/hour. The reducing agent was 0.3 mM STS and 3 ml was added with 3 ml/hour after 7.5 hours after the preparation procedure was started.
fluid bed dryer for 60 minutes at 60°C and for several minutes at 90 °C (L = 15 minutes; C = 30 minutes and H = 45 minutes) and immediately added to a rotating reactor with a constant temperature of 160°C. The product was removed from the reactor at x hours after addition (L = 3 hours; C = 5 hours and H = 7 hours) and stored at room temperature in a plastic bag after cooling down on a piece of paper for 2 hours.
Experimental designA Taguchi L8 orthogonal design allows the evaluation of the three factors (i.e. HCl, PDM and DTH) and their interactions [9]. The effect of a factor, or an interaction, is in the simplified Taguchi method calculated by multiplying the measured value with the corresponding number in the Taguchi L8 matrix (Table 1). The effect of a factor, or an interaction, is defined as the sum of the obtained values of the eight runs performed after this multiplication. The Taguchi L8 design was augmented with a centre point (C). This setting was executed in triplicate in order to get an impression of the overall level of noise in the pyrodextrination procedure [9,33].
Table 1: Taguchi L8 matrix for three factors and their interactions.Run Code Factors or interactions
AT BT ATBT CT ATCT BTCT ATBTCT
1 LLL -1 -1 -1 -1 -1 -1 -12 LLH -1 -1 -1 1 1 1 13 LHL -1 1 1 -1 -1 1 14 LHH -1 1 1 1 1 -1 -15 HLL 1 -1 1 -1 1 -1 16 HLH 1 -1 1 1 -1 1 -17 HHL 1 1 -1 -1 1 1 -18 HHH 1 1 -1 1 -1 -1 1
Polymerization equipmentA double-jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles) with a width of 1 cm and an outer dimension of 10.5x7 cm (height x diameter)) was used. A lid made of borosilicate glass with several connection points was placed on top and the reactor was completely insulated with radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the temperature of the headspace in the reactor (HST). The feeding lines of VAM and the SPS/SBC mixture were placed outside the reflux region with the aid of an accessory to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 4).
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
58 59
dispersion in 100 ml acetone (0-5 °C) followed by soxhlet extraction in acetone. The soxhlet extraction was automated with a Soxtec 2043/2046 system from Foss Analytical (Cellulose thimbles, 160°C; 6 hours boiling and 18 hours rinsing). The grafting efficiency (f) is defined as:
Equation 3:
Where , and denote the amount of grafted, total and homo-polymerized VAM. The grafted copolymer amount was simply taken as the dried residue after the acetone extraction. The total pVAc was calculated on the basis of the monomer intake and conversion values.Glass transition temperatures (Tg) were derived from total heat flow and reversing heat flow curves determined with a modulated differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg dispersion was added to a pan without any preliminary handling steps).
ResultsPyrodextrin preparationA Taguchi procedure was applied on the variables Ostwald viscosity, YI-E313 and CIE L*a*b*. The results of the CCC-pyrodextrin triplicate were used to estimate the level of variation. The effects of the three selected factors (HCl, PDM and DHT), and their interactions, which are larger than two times the standard deviation (σ) of the CCC, are shown in Table 2 (the original data is not shown for brevity).The effect of the triple interaction is frequently used as an estimator of the level of noise in this type of experiments and therefore not evaluated. The reaction conditions for pyrodextrination had a pronounced impact on the (macro)molecular size of these products as testified by the relatively broad range of the Ostwald viscosity values (from 37 to 78 mPa·s). The corresponding modification mechanisms are controlled, according the simplified Taguchi calculations, by HCl, DTH and their interaction. These two factors control the acidic breakdown and transglycosylation [11-15]. This suggests that the pyrodextrination process is mainly controlled by these two molecular processes. Pronounced shifts in colouration (i.e. YI-E313, CIE L*, CIE a* or CIE b*) were observed as function of HCl and DTH and the impact of their interaction was smaller and in the opposite direction. The obtained results (i.e. magnitude and sign of the observed effects in Table 2) are in line with the fact that the viscosity and colouration are controlled by different processes. The variation in heat exposure introduced by the PDM procedure appears to be negligible with respect to DTH. The effect of PDM was not significant according the simplified Taguchi calculations and the interaction between DTH and PDM barely exceeds the level of noise calculated with the CCC setting. YI-E313 correlates with CIE b* (based on the sign of the coefficients in Table 2, columns 3 and 6) as expected because this parameter describes the blue - yellow shift of the colour analysed. CIE L* and a* show that the impact of the applied changes was not limited to the degree of yellowness only. The observed differences in colouration suggest that the applied settings did not only control the number of colouring groups generated during processing, but the type of colouring groups formed as well.
Figure 5: Applied temperature profile and dosage protocols of VAM, SPS/SBC and STS.
Characterization of pyrodextrins and corresponding latexesA Datacolor Spectraflash SF-450 was used to determine the colour of the pyrodextrins after complete dissolution in water (concentration 4 wt %; demi water; settings: illuminator, D65; 10°; spectrum, 360–700 nm; mode, transmission/absorption). The Ostwald viscosity was determined by measurement of a solution with 58% refraction at 25°C. The WPC is expressed in a percentage in the range of -100 to 100 % and this value was recorded every 3 seconds during the preparation procedure. The amount of water present in the water bath was approximately the same in each experiment in order to minimize the variation between the executed experiments. The average WPC of 0-0.5 hours was calculated and defined as the heat loss to the environment. This value was subtracted from all recorded values before the sum was taken from all the readings in the period from 0-7 hours. The obtained value was used as a measure for the WPC during polymerizationViscosity, pH and dry matter of the obtained latexes were determined with the help of a Brookfield DV-II+( 20RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80 °C), respectively. Ethanal and residual VAM in the latexes were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) and a flame ionization detector. The gas chromatography measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 minutes and the supernatant was mixed 1:1 with 5 mM NaOH in water. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac colomns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions (sulfate, thiosulfate and free acetate) was achieved with a gradient of sodium hydroxide. PSD’s were obtained with a Sympatec laser diffractor (LD) equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13-32.5 μm). The sample solution was circulated during measurement. Fraunhofer theory based calculations were used and the obtained particle size distributions are ISO 13320 compliant. Dynamic Light Scattering (DLS) measurements were performed with a Zetapals from Brookhaven (4 ml demi water and 2 drops of diluted latex (10 wt % in demi water); angle 90°; average of 10 measurements of 30 seconds). The amount of grafted (and/or complex mixture) material was determined based on a precipitation of 3-4 g
0
5
10
15
20
25
0
20
40
60
80
100
-1 0 1 2 3 4 5 6 7 8 9
Mono
mer (
mmol/
min)
Temp
eratu
re (°
C)
Time (hours) Profile: Water bath Dosage: VAM
SPS/SBC: 5 ml (5 ml/min) + 25.5 ml (5.8 ml/hr) STS: 3 ml (3 ml/hr)
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
58 59
dispersion in 100 ml acetone (0-5 °C) followed by soxhlet extraction in acetone. The soxhlet extraction was automated with a Soxtec 2043/2046 system from Foss Analytical (Cellulose thimbles, 160°C; 6 hours boiling and 18 hours rinsing). The grafting efficiency (f) is defined as:
Equation 3:
Where , and denote the amount of grafted, total and homo-polymerized VAM. The grafted copolymer amount was simply taken as the dried residue after the acetone extraction. The total pVAc was calculated on the basis of the monomer intake and conversion values.Glass transition temperatures (Tg) were derived from total heat flow and reversing heat flow curves determined with a modulated differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg dispersion was added to a pan without any preliminary handling steps).
ResultsPyrodextrin preparationA Taguchi procedure was applied on the variables Ostwald viscosity, YI-E313 and CIE L*a*b*. The results of the CCC-pyrodextrin triplicate were used to estimate the level of variation. The effects of the three selected factors (HCl, PDM and DHT), and their interactions, which are larger than two times the standard deviation (σ) of the CCC, are shown in Table 2 (the original data is not shown for brevity).The effect of the triple interaction is frequently used as an estimator of the level of noise in this type of experiments and therefore not evaluated. The reaction conditions for pyrodextrination had a pronounced impact on the (macro)molecular size of these products as testified by the relatively broad range of the Ostwald viscosity values (from 37 to 78 mPa·s). The corresponding modification mechanisms are controlled, according the simplified Taguchi calculations, by HCl, DTH and their interaction. These two factors control the acidic breakdown and transglycosylation [11-15]. This suggests that the pyrodextrination process is mainly controlled by these two molecular processes. Pronounced shifts in colouration (i.e. YI-E313, CIE L*, CIE a* or CIE b*) were observed as function of HCl and DTH and the impact of their interaction was smaller and in the opposite direction. The obtained results (i.e. magnitude and sign of the observed effects in Table 2) are in line with the fact that the viscosity and colouration are controlled by different processes. The variation in heat exposure introduced by the PDM procedure appears to be negligible with respect to DTH. The effect of PDM was not significant according the simplified Taguchi calculations and the interaction between DTH and PDM barely exceeds the level of noise calculated with the CCC setting. YI-E313 correlates with CIE b* (based on the sign of the coefficients in Table 2, columns 3 and 6) as expected because this parameter describes the blue - yellow shift of the colour analysed. CIE L* and a* show that the impact of the applied changes was not limited to the degree of yellowness only. The observed differences in colouration suggest that the applied settings did not only control the number of colouring groups generated during processing, but the type of colouring groups formed as well.
Figure 5: Applied temperature profile and dosage protocols of VAM, SPS/SBC and STS.
Characterization of pyrodextrins and corresponding latexesA Datacolor Spectraflash SF-450 was used to determine the colour of the pyrodextrins after complete dissolution in water (concentration 4 wt %; demi water; settings: illuminator, D65; 10°; spectrum, 360–700 nm; mode, transmission/absorption). The Ostwald viscosity was determined by measurement of a solution with 58% refraction at 25°C. The WPC is expressed in a percentage in the range of -100 to 100 % and this value was recorded every 3 seconds during the preparation procedure. The amount of water present in the water bath was approximately the same in each experiment in order to minimize the variation between the executed experiments. The average WPC of 0-0.5 hours was calculated and defined as the heat loss to the environment. This value was subtracted from all recorded values before the sum was taken from all the readings in the period from 0-7 hours. The obtained value was used as a measure for the WPC during polymerizationViscosity, pH and dry matter of the obtained latexes were determined with the help of a Brookfield DV-II+( 20RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80 °C), respectively. Ethanal and residual VAM in the latexes were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) and a flame ionization detector. The gas chromatography measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 minutes and the supernatant was mixed 1:1 with 5 mM NaOH in water. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac colomns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions (sulfate, thiosulfate and free acetate) was achieved with a gradient of sodium hydroxide. PSD’s were obtained with a Sympatec laser diffractor (LD) equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13-32.5 μm). The sample solution was circulated during measurement. Fraunhofer theory based calculations were used and the obtained particle size distributions are ISO 13320 compliant. Dynamic Light Scattering (DLS) measurements were performed with a Zetapals from Brookhaven (4 ml demi water and 2 drops of diluted latex (10 wt % in demi water); angle 90°; average of 10 measurements of 30 seconds). The amount of grafted (and/or complex mixture) material was determined based on a precipitation of 3-4 g
0
5
10
15
20
25
0
20
40
60
80
100
-1 0 1 2 3 4 5 6 7 8 9
Mono
mer (
mmol/
min)
Temp
eratu
re (°
C)
Time (hours) Profile: Water bath Dosage: VAM
SPS/SBC: 5 ml (5 ml/min) + 25.5 ml (5.8 ml/hr) STS: 3 ml (3 ml/hr)
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
60 61
Figure 7: WPC profiles of the VAM polymerizations with the prepared pyrodextrins as protective colloid. The error bars represent the two times σ of the triplicate.
Simplified Taguchi L8 calculations were applied on the variables total WPC during polymerization, recovery, pH, residual ethanal, remnant VAM and amount of sulfate and thiosulfate ions. Two times σ of the CCC pyrodextrin protected latexes was used as a threshold value. The results of these calculations that exceed both thresholds are given in Table 3 (original values are not shown for brevity).
Table 3: Effect of the different pyrodextrins on the selected polymerization variables based on Taguchi calculations (simplified). Only effects are shown which exceed two times the σ of the triplicate of the CCC pyrodextrins.
Factors WPC Latex composition
Sum 0-7 hours (%)
Ethanal(mg/g)
VAM(mg/g)
Dry matter
(%)pH Acetate
(mmol)Sulfate(mmol)
Thiosulfate(mmol)
PDM - - - - - - - -HCl 31 -0.3 2.8 -0.8 0.07 2.57 - 0.04PDM:HCl - - - - - - - -DTH 26 - 1.6 -0.8 0.06 1.84 - 0.06DTH:PDM - - - - - - - -DTH:HCl -28 - -0.7 0.3 - - - 0.06DTH:PDM:HCl 27 - - - - - - -0.05CCC (Average) 131 1.4 6.0 98.6 4.97 36.02 3.33 0.34CCC (2*σ) 19 0.2 0.4 0.1 0.02 1.44 0.14 0.02
The triple interactions were excluded from this evaluation. The total WPC during polymerization was affected in a significant way by the factors HCl and DTH. Both factors have, on their own and in combination, a pronounced influence on the variables residual vinyl acetate and YI-E313 (10.1, 11.7 and -1.9 respectively; Table 3). The total recovery of the latex was also distinctly influenced by the factors HCl and DTH and their interaction. The factors HCl and DTH need to be minimized for a quick and optimal polymerization process in general. The factor HCl did have a slight negative effect on the amount of residual ethanal of the latexes.
20
30
40
50
60
0 1 2 3 4 5 6
Pow
er c
onsu
mpt
ion
(%)
Time (Hours)
Average (CCC) HHH LHH
Table 2: Effect of different reaction conditions on the selected pyrodextrin variables based on Taguchi calculations (simplified). Only effects are shown which exceed two times the standard deviation (σ) of the triplicate of the CCC pyrodextrin.
Factors Viscosity Dextrin characteristicsOstwald (mPa·s) YI-E313 CIE L* CIE a* CIE b*
PDM - - - - -HCl -26.3 10.1 -1.1 -1.2 6.3PDM:HCl - - - - -DTH -13.3 11.7 -0.9 -1.1 7.2DTH:PDM - 1.4 -0.3 - 0.9DTH:HCl -11.8 -1.9 0.5 - -1.3DTH:PDM:HCl - - - - -CCC (Average) 44.0 17.4 94.5 -2.0 10.3CCC (2*σ) 0.4 1.1 0.1 0.1 0.7
Polymerization process characteristicsThe selected VAM polymerization was executed with the CCC pyrodextrin and the profiles HST and WPC were plotted against time (Figure 6). The drop in HST to 66 °C at 0.5 hours was accompanied with an increase in the WPC. The sharp rise in the HST at ~4.5 hours coincided with a reduction in the WPC. This is indicative for a considerable level of refluxing of the binary azeotrope VAM and water mainly during VAM dosage. However, there is a considerable chance that azeotrope is contaminated with other ingredients of the reaction mixture (e.g. ethanal and acetic acid) [9]. The WPC was approximately 40 % during refluxing which was distinctly higher than the ~30 % when there was no VAM dosed and considerably lower than the maximum WPC possible. These characteristics offer a good starting point for evaluating the impact of different pyrodextrins on the WPC profile.Polymerizations with the two pyrodextrins with highest degree of yellowness resulted in distinct WPC’s (Figure 7). The obtained results were indicative for differences in level of refluxing and the composition of the products with YI-E313 ≈ 30 results apparently in a significant change in polymerization rate.
Figure 6: HST and WPC profiles of a VAM polymerization with the CCC pyrodextrin as protective colloid.
0
20
40
60
80
100
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9
Pow
er c
onsu
mpt
ion
(%)
Ref
lux
Tem
pera
ture
(°C
)
Time (Hours)
HST WPC
VAM: End VAM: Start
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
60 61
Figure 7: WPC profiles of the VAM polymerizations with the prepared pyrodextrins as protective colloid. The error bars represent the two times σ of the triplicate.
Simplified Taguchi L8 calculations were applied on the variables total WPC during polymerization, recovery, pH, residual ethanal, remnant VAM and amount of sulfate and thiosulfate ions. Two times σ of the CCC pyrodextrin protected latexes was used as a threshold value. The results of these calculations that exceed both thresholds are given in Table 3 (original values are not shown for brevity).
Table 3: Effect of the different pyrodextrins on the selected polymerization variables based on Taguchi calculations (simplified). Only effects are shown which exceed two times the σ of the triplicate of the CCC pyrodextrins.
Factors WPC Latex composition
Sum 0-7 hours (%)
Ethanal(mg/g)
VAM(mg/g)
Dry matter
(%)pH Acetate
(mmol)Sulfate(mmol)
Thiosulfate(mmol)
PDM - - - - - - - -HCl 31 -0.3 2.8 -0.8 0.07 2.57 - 0.04PDM:HCl - - - - - - - -DTH 26 - 1.6 -0.8 0.06 1.84 - 0.06DTH:PDM - - - - - - - -DTH:HCl -28 - -0.7 0.3 - - - 0.06DTH:PDM:HCl 27 - - - - - - -0.05CCC (Average) 131 1.4 6.0 98.6 4.97 36.02 3.33 0.34CCC (2*σ) 19 0.2 0.4 0.1 0.02 1.44 0.14 0.02
The triple interactions were excluded from this evaluation. The total WPC during polymerization was affected in a significant way by the factors HCl and DTH. Both factors have, on their own and in combination, a pronounced influence on the variables residual vinyl acetate and YI-E313 (10.1, 11.7 and -1.9 respectively; Table 3). The total recovery of the latex was also distinctly influenced by the factors HCl and DTH and their interaction. The factors HCl and DTH need to be minimized for a quick and optimal polymerization process in general. The factor HCl did have a slight negative effect on the amount of residual ethanal of the latexes.
20
30
40
50
60
0 1 2 3 4 5 6
Pow
er c
onsu
mpt
ion
(%)
Time (Hours)
Average (CCC) HHH LHH
Table 2: Effect of different reaction conditions on the selected pyrodextrin variables based on Taguchi calculations (simplified). Only effects are shown which exceed two times the standard deviation (σ) of the triplicate of the CCC pyrodextrin.
Factors Viscosity Dextrin characteristicsOstwald (mPa·s) YI-E313 CIE L* CIE a* CIE b*
PDM - - - - -HCl -26.3 10.1 -1.1 -1.2 6.3PDM:HCl - - - - -DTH -13.3 11.7 -0.9 -1.1 7.2DTH:PDM - 1.4 -0.3 - 0.9DTH:HCl -11.8 -1.9 0.5 - -1.3DTH:PDM:HCl - - - - -CCC (Average) 44.0 17.4 94.5 -2.0 10.3CCC (2*σ) 0.4 1.1 0.1 0.1 0.7
Polymerization process characteristicsThe selected VAM polymerization was executed with the CCC pyrodextrin and the profiles HST and WPC were plotted against time (Figure 6). The drop in HST to 66 °C at 0.5 hours was accompanied with an increase in the WPC. The sharp rise in the HST at ~4.5 hours coincided with a reduction in the WPC. This is indicative for a considerable level of refluxing of the binary azeotrope VAM and water mainly during VAM dosage. However, there is a considerable chance that azeotrope is contaminated with other ingredients of the reaction mixture (e.g. ethanal and acetic acid) [9]. The WPC was approximately 40 % during refluxing which was distinctly higher than the ~30 % when there was no VAM dosed and considerably lower than the maximum WPC possible. These characteristics offer a good starting point for evaluating the impact of different pyrodextrins on the WPC profile.Polymerizations with the two pyrodextrins with highest degree of yellowness resulted in distinct WPC’s (Figure 7). The obtained results were indicative for differences in level of refluxing and the composition of the products with YI-E313 ≈ 30 results apparently in a significant change in polymerization rate.
Figure 6: HST and WPC profiles of a VAM polymerization with the CCC pyrodextrin as protective colloid.
0
20
40
60
80
100
0
20
40
60
80
100
0 1 2 3 4 5 6 7 8 9
Pow
er c
onsu
mpt
ion
(%)
Ref
lux
Tem
pera
ture
(°C
)
Time (Hours)
HST WPC
VAM: End VAM: Start
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
62 63
0.0
0.5
1.0
1.5
2.0
0.1 1 10 100
Dens
ity
Size (µm)
Average (CCC) HHH LHH
Figure 8: PSD’s of a number of characteristic pyrodextrin based latexes. The error bars represent the two times σ of the triplicate.
Protective colloid related propertiesThe values for soluble and insoluble fractions from the acetone precipitation and extraction are given in Table A2 together with the calculated amounts of homopolymer, grafted polymer and f. Table 4 shows the Taguchi calculations of the selected pyrodextrination conditions, and their interactions, which are larger than two times the standard deviation of the corresponding CCC pyrodextrin. An increase in HCl during the pyrodextrination resulted in a higher f for the final latex, which can be partly attributed to differences in the HCl content of the pyrodextrins prepared. However, the HCl content in the pyrodextrin is only 0.1 wt % and the influence of variations in this area can therefore be ignored. The interaction between the factors DTH and HCl on the protective colloid properties of the corresponding pyrodextrins is remarkable. These latexes show a reduced level of grafting efficiency with respect to their counterparts based on CCC pyrodextrins but the level of monodispersity of the corresponding PSD’s were considerably improved (Figure 8). Latex particles were more stabilized with this type of pyrodextrins apparently. Stabilization by a higher viscosity level of the water phase present during processing can be ruled out because the interaction of HCl and DHT resulted in a decrease in viscosity. The interaction also resulted in a lower degree of yellowing but the actual impact was small (Table 2). The effect of the different pyrodextrins is not limited to the characteristics of the final latex only. There were distinct differences between the WPC profiles between polymerizations based on the pyrodextrins HHH and LHH and the other varieties (Figure 7), which was in turn strongly related to level of reflux of the azeotrope VAM and water. Pyrodextrins HHH and LHH might therefore induce significantly different reaction conditions during polymerization with respect to reaction temperature and amount of water (temporarily) withdrawn from the reaction mixture due to refluxing. Against this backdrop, the presence of several (statistically
The differences in WPC during processing and residual VAM of the latex might be indicative for changes in the radical formation process. Unfortunately, the observed differences in pH and anion composition of the latexes were small and did not provide any confirming evidence. The observed variations in residual STS were close to the level of the triple interaction and therefore considered to be negligible. The amount of SPS at the moment that the STS solution was dosed was therefore comparable in all cases and the amount of sulfate generated during processing was close to the same as well. Residual VAM and YI-E313 appear to be proportionally linked and this suggests that the pyrodextrin colouration had a distinct influence on the radical formation process. This assumption is plausible because an increase in YI-E313 is associated with an increase in conjugated double bonds and these types of bonds are known as radical scavengers [23]. However, chloride is a radical inhibitor as well and the fact that the chloride content of the pyrodextrins used is not the same in all cases makes deriving solid conclusions difficult [28,29]. The radical formation process is therefore influenced by either chloride content, colouration or a combination thereof. Differences were observed for pH and concentration acetate but they were small and can probably be related to the composition of the pyrodextrin and the amount of residual VAM respectively. From this it can then be concluded that HCl and DTH are essential for further optimization.
Latex characteristicsThe factors DTH and HCl appear to be crucial for creating a pyrodextrin that is able to make latex with a monodisperse PSD (Figure 8). Only pyrodextrins prepared at the settings HHH and LHH generated PSD’s that display a certain degree of monodispersity. The PSD’s of the other latexes were similar to the PSD of the pyrodextrin prepared at the CCC setting and this latex was not monodisperse at all (Figure 8). The calculated variables from the above mentioned LD based PSD’s were not suitable for Taguchi calculations. Therefore, DLS measurements at a 90° angle were used to determine the PSD of the latexes below 1 μm. The results are displayed in Table A1 together with the values of the latex viscosity and the Tg’s based on Onset (Tg,onset ), Inflection (Tg,inflection ) and Endset (Tg,endset). However, Taguchi calculations (not shown for brevity) show that these three latex variables were barely influenced by the type of pyrodextrins used. These considerations are on one side very positive since they clearly indicate the robustness of the process in the range of experimental conditions investigated. On the other hand, the lack of a strong correlation does not allow, at least for the thermal properties, identifying the experimental variables that can be used for further optimization.
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
62 63
0.0
0.5
1.0
1.5
2.0
0.1 1 10 100
Dens
ity
Size (µm)
Average (CCC) HHH LHH
Figure 8: PSD’s of a number of characteristic pyrodextrin based latexes. The error bars represent the two times σ of the triplicate.
Protective colloid related propertiesThe values for soluble and insoluble fractions from the acetone precipitation and extraction are given in Table A2 together with the calculated amounts of homopolymer, grafted polymer and f. Table 4 shows the Taguchi calculations of the selected pyrodextrination conditions, and their interactions, which are larger than two times the standard deviation of the corresponding CCC pyrodextrin. An increase in HCl during the pyrodextrination resulted in a higher f for the final latex, which can be partly attributed to differences in the HCl content of the pyrodextrins prepared. However, the HCl content in the pyrodextrin is only 0.1 wt % and the influence of variations in this area can therefore be ignored. The interaction between the factors DTH and HCl on the protective colloid properties of the corresponding pyrodextrins is remarkable. These latexes show a reduced level of grafting efficiency with respect to their counterparts based on CCC pyrodextrins but the level of monodispersity of the corresponding PSD’s were considerably improved (Figure 8). Latex particles were more stabilized with this type of pyrodextrins apparently. Stabilization by a higher viscosity level of the water phase present during processing can be ruled out because the interaction of HCl and DHT resulted in a decrease in viscosity. The interaction also resulted in a lower degree of yellowing but the actual impact was small (Table 2). The effect of the different pyrodextrins is not limited to the characteristics of the final latex only. There were distinct differences between the WPC profiles between polymerizations based on the pyrodextrins HHH and LHH and the other varieties (Figure 7), which was in turn strongly related to level of reflux of the azeotrope VAM and water. Pyrodextrins HHH and LHH might therefore induce significantly different reaction conditions during polymerization with respect to reaction temperature and amount of water (temporarily) withdrawn from the reaction mixture due to refluxing. Against this backdrop, the presence of several (statistically
The differences in WPC during processing and residual VAM of the latex might be indicative for changes in the radical formation process. Unfortunately, the observed differences in pH and anion composition of the latexes were small and did not provide any confirming evidence. The observed variations in residual STS were close to the level of the triple interaction and therefore considered to be negligible. The amount of SPS at the moment that the STS solution was dosed was therefore comparable in all cases and the amount of sulfate generated during processing was close to the same as well. Residual VAM and YI-E313 appear to be proportionally linked and this suggests that the pyrodextrin colouration had a distinct influence on the radical formation process. This assumption is plausible because an increase in YI-E313 is associated with an increase in conjugated double bonds and these types of bonds are known as radical scavengers [23]. However, chloride is a radical inhibitor as well and the fact that the chloride content of the pyrodextrins used is not the same in all cases makes deriving solid conclusions difficult [28,29]. The radical formation process is therefore influenced by either chloride content, colouration or a combination thereof. Differences were observed for pH and concentration acetate but they were small and can probably be related to the composition of the pyrodextrin and the amount of residual VAM respectively. From this it can then be concluded that HCl and DTH are essential for further optimization.
Latex characteristicsThe factors DTH and HCl appear to be crucial for creating a pyrodextrin that is able to make latex with a monodisperse PSD (Figure 8). Only pyrodextrins prepared at the settings HHH and LHH generated PSD’s that display a certain degree of monodispersity. The PSD’s of the other latexes were similar to the PSD of the pyrodextrin prepared at the CCC setting and this latex was not monodisperse at all (Figure 8). The calculated variables from the above mentioned LD based PSD’s were not suitable for Taguchi calculations. Therefore, DLS measurements at a 90° angle were used to determine the PSD of the latexes below 1 μm. The results are displayed in Table A1 together with the values of the latex viscosity and the Tg’s based on Onset (Tg,onset ), Inflection (Tg,inflection ) and Endset (Tg,endset). However, Taguchi calculations (not shown for brevity) show that these three latex variables were barely influenced by the type of pyrodextrins used. These considerations are on one side very positive since they clearly indicate the robustness of the process in the range of experimental conditions investigated. On the other hand, the lack of a strong correlation does not allow, at least for the thermal properties, identifying the experimental variables that can be used for further optimization.
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
64 65
AcknowledgementsThis investigation was sponsored by Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
AbbreviationspVAc : Polyvinyl acetate.HMF : Hydroxy methyl fufural.YI-400nm: Yellow Index based on 400nm.[η] : Intrinsic viscosity.YI-E313 : Yellow Index based on CIE E313.CIE : Commission Internationale de l’Eclairage.CIE L* : Black - white shift or intensity (colour space)CIE a* : Green - red shift (colour space) CIE b* : Blue - yellow shift (colour space)HCl : Hydrochloric acid.PDM : Pre-drying in minutes.DTH : Dextrination time in hoursVAM : Vinyl acetate monomer.SPS: Sodium persulfate.SBC: Sodium bicarbonate.STS: Sodium thiosulfate.HST: Headspace temperature.WPC: Water bath power consumption.PSD: Particle size distribution.L : Low level.H : High level.C : Centre level.AT : PDM.BT : HCl.CT : DTH.RPM : Revolutions per minute of stirrer.LD: Laser diffractor.DLS: Dynamic light scattering.f : Grafting efficiency. : Amount of grafted pVAc. : Amount of total pVAc. : Amount of homopolymer pVAc.Tg : Glass transition temperature.mDSC: Modulated diffential scanning calorimeter.σ : Standard deviation.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
significant) interaction coefficients between the different factors (Table 4) clearly indicated the complexity of the grafting reaction mechanism. These considerations render the use of a statistical procedure quite attractive for optimization at industrial level, but they also invite the thought (beyond the scope of the present work) of a more detailed study on the molecular mechanism.
Table 4: Effect of different reaction conditions on the selected pVAc composition variables (Table A2) based on Taguchi calculations (simplified). Only effects are shown which exceed two times σ of the triplicate of the CCC pyrodextrin.
Factors pVAc compositionHomopolymer
(mg/g)Grafted polymer
(mg/g)f
(fraction)PDM - - -HCl -55 57 0.15PDM:HCl - - 0.06DTH -25 24 0.07DTH:PDM - - -0.03DTH:HCl 45 -45 -0.12DTH:PDM:HCl - - 0.04CCC: Average 269 96 0.24CCC: 2*σ 24 23 0.02
ConclusionsThe hydrochloric acid (HCl) content and the heat exposure time (DTH) constitute two crucial factors controlling not only the colour and viscosity of pyrodextrins, but also the protective colloid properties of these products. Heat flow during processing, monomer conversion, product recovery, anion concentration, pH, viscosity, particle size distribution, amount of grafted protective colloid and level of viscosity show distinct changes depending on the pyrodextrine used whilst the glass transition temperature remains approximately the same. The water bath power consumption during the latex preparation is proportional to the amount of HCl used during the pyrodextrination process, thus providing in principle a simple strategy to minimize the energy input for the latex synthesis. However, a high energy consuming polymerization procedure appears to be a necessity for creating a monodisperse particle size distributions. The performed statistical analysis is useful in identifying the interplay between the different investigated factors, indicating in some cases (e.g. grafting mechanism on molecular scale) the necessity of more dedicated studies. The selected reaction conditions for the pyrodextrination process do not reveal any straightforward strategies to create a latex with a monodisperse particle size distribution by reducing the amount of HCl used in the pyrodextrination. In this context, a significant change in pyrodextrination process and/or polymerization conditions is needed if this drawback needs to be solved. On the other hand, the employed statistical analysis reveals clearly the intimate relationship between the pyrodextrins synthesis and the properties of the polyvinyl acetate latex obtained by using them as protective colloid. To the best of our knowledge this represents a relevant novelty of the present work, clearly suggesting the need for more accurate macromolecular characterization if one wishes to establish a reliable structure-property relationship.
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
64 65
AcknowledgementsThis investigation was sponsored by Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
AbbreviationspVAc : Polyvinyl acetate.HMF : Hydroxy methyl fufural.YI-400nm: Yellow Index based on 400nm.[η] : Intrinsic viscosity.YI-E313 : Yellow Index based on CIE E313.CIE : Commission Internationale de l’Eclairage.CIE L* : Black - white shift or intensity (colour space)CIE a* : Green - red shift (colour space) CIE b* : Blue - yellow shift (colour space)HCl : Hydrochloric acid.PDM : Pre-drying in minutes.DTH : Dextrination time in hoursVAM : Vinyl acetate monomer.SPS: Sodium persulfate.SBC: Sodium bicarbonate.STS: Sodium thiosulfate.HST: Headspace temperature.WPC: Water bath power consumption.PSD: Particle size distribution.L : Low level.H : High level.C : Centre level.AT : PDM.BT : HCl.CT : DTH.RPM : Revolutions per minute of stirrer.LD: Laser diffractor.DLS: Dynamic light scattering.f : Grafting efficiency. : Amount of grafted pVAc. : Amount of total pVAc. : Amount of homopolymer pVAc.Tg : Glass transition temperature.mDSC: Modulated diffential scanning calorimeter.σ : Standard deviation.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
significant) interaction coefficients between the different factors (Table 4) clearly indicated the complexity of the grafting reaction mechanism. These considerations render the use of a statistical procedure quite attractive for optimization at industrial level, but they also invite the thought (beyond the scope of the present work) of a more detailed study on the molecular mechanism.
Table 4: Effect of different reaction conditions on the selected pVAc composition variables (Table A2) based on Taguchi calculations (simplified). Only effects are shown which exceed two times σ of the triplicate of the CCC pyrodextrin.
Factors pVAc compositionHomopolymer
(mg/g)Grafted polymer
(mg/g)f
(fraction)PDM - - -HCl -55 57 0.15PDM:HCl - - 0.06DTH -25 24 0.07DTH:PDM - - -0.03DTH:HCl 45 -45 -0.12DTH:PDM:HCl - - 0.04CCC: Average 269 96 0.24CCC: 2*σ 24 23 0.02
ConclusionsThe hydrochloric acid (HCl) content and the heat exposure time (DTH) constitute two crucial factors controlling not only the colour and viscosity of pyrodextrins, but also the protective colloid properties of these products. Heat flow during processing, monomer conversion, product recovery, anion concentration, pH, viscosity, particle size distribution, amount of grafted protective colloid and level of viscosity show distinct changes depending on the pyrodextrine used whilst the glass transition temperature remains approximately the same. The water bath power consumption during the latex preparation is proportional to the amount of HCl used during the pyrodextrination process, thus providing in principle a simple strategy to minimize the energy input for the latex synthesis. However, a high energy consuming polymerization procedure appears to be a necessity for creating a monodisperse particle size distributions. The performed statistical analysis is useful in identifying the interplay between the different investigated factors, indicating in some cases (e.g. grafting mechanism on molecular scale) the necessity of more dedicated studies. The selected reaction conditions for the pyrodextrination process do not reveal any straightforward strategies to create a latex with a monodisperse particle size distribution by reducing the amount of HCl used in the pyrodextrination. In this context, a significant change in pyrodextrination process and/or polymerization conditions is needed if this drawback needs to be solved. On the other hand, the employed statistical analysis reveals clearly the intimate relationship between the pyrodextrins synthesis and the properties of the polyvinyl acetate latex obtained by using them as protective colloid. To the best of our knowledge this represents a relevant novelty of the present work, clearly suggesting the need for more accurate macromolecular characterization if one wishes to establish a reliable structure-property relationship.
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
66 67
(http://www.machinerylubrication.com/Read/920/kinematic-viscosity-measurement; 20-06-2012)[28] L.R. Bennedsen, J. Muff, E.G. Sogaard, Influence of chloride and carbonates on the reactivity of activated
persulfate, Chemosphere, 86, 11 (2012) 1092-1097[29] M.Hahn, W. Jaeger, C. Wandrey, G. Reinisch, Zur kinetik der radikalischen polymerization von dimethyl-
diallyl-ammoniumchloride: IV/ Mechanismus von start- und abbruchreaction met persulfat als initiator, Acta Polymerica 35 (5) (1984) 350-358
[30] J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical engineering: Green engineering, Chemical Engineering Journal, 133 (2007) 7-30
[31] ACS Green Chemistry Institute, The twelve principles of green engineering,( www.acs.org; 18-03-2013)[32] ACS Green Chemistry Institute, The twelve principles of green chemistry (www.acs.org; 18-03-2013)[33] Experimental design and Taguchi (http://homepage.ntlworld.com/s.orszulik/index.html; 20-06-2012)
References[1] I. Skeist, Handbook of adhesives, Reinhold Publishing Corp. Chapman & Hall, London, 1962[2] V.A. Lauria, Remoistenable adhesive compositions (1987) US4678824[3] M.S. Mahiel, J.M. Cruden, Adhesive compositions and self-adhesive sheet materials (1988) EP0297900A2[4] M.S. Mahiel, J.M. Cruden, Surface coating compositions (1989) EP0351193A2[5] Jr.J. Wieczorek, L.M. Mahony, Aqueous adhesive compositions for use in binding book (1996) US5519072[6] U. Geissler, H. Hintz, U. Vogt-saggau, Powdery adhesive composition (1997) EP0799876A2[7] T. Mayer, H.P. Weitzel, R. Haerschel, T. Bastelberger, Method for producing polymers stabilized with
protective colloids (2000) US6300403[8] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum
adhesive based on a filled polymer dispersion, WO2006094594A1.[9] K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman,
H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects. To be published.
[10] A.A.M. Maas, J.C.P. Hopman, R.P.W. Kesselmans, Dextrinization of starch (2003) US6613152[11] B. Brimhall, Structure of pyrodextrins, Industrial and engineering chemistry, 36 (1) (1944) 72-75[12] R.J. Dimler, H.A. Davis, G.E. Hilbert, A new anhydride of D-glucose: D-Glucosan ,<1,4> β <1,6>, Journal of
the american chemical society, 68 (7) (1946) 1377-1380[13] R.W. Kerr, F.C. Cleveland, Chemistry of dextrination, Starch, 5 (10) (1953) 261-266[14] M.L. Wolfrom, A. Thompson, R.B. Ward, Composition of pyrodextrins, Industrial and engineering chemistry,
55 (3) (1961) 217-28[15] D.J. Bryce, C.T. Greenwood, Aspects of thermal degradation of starch, Starch, 15 (5) (1963) 166-170[16] A.A. Rosatella, S.P. Simoneonov, R.F.M. Frade, C.A.M. Alfonso, 5-Hydroxymethylfurfural (HMF) as a
building block platform: Biological properties, synthesis and synthetic applications, Green Chemistry, 13 (2011) 754-793
[17] S.K.R. Patil, C.R.F. Lund, Formation and Growth of Humins via Aldol Addition and Condensation during Acid-Catalyzed Conversion of 5-Hydroxymethylfurfural, Energy and Fuels, 25 (2011) 4745-4755
[18] H.E. van Dam, H.P.G. Kieboom,H van Bekkum, The conversion of fructose and glucose in acidic media: Formation of hydroxymethylfurfural, 38 (3) (1986) 95-101
[19] L.A. Ameur, O. Mathieu, V. Lalanne, G. Trystram, I. Birlouez-Aragon, Comparison of the effects of sucrose and hexose on furfural formation and browning in cookies baked at different temperatures. Food Chemistry 101 (2007) 1407–1416.
[20] K. R. Terpstra, A. J. J. Woortman, J. C. P. Hopman, Yellow dextrins: Evaluating changes in structure and colour during processing, Starch, 62 (9) (2010) 449-457
[21] R.J. van Putten, J.C. van der Waal, E. de Jong, C.B. Rasendra, H.J. Heeres, J.G. de Vries, Hydroxymethylfurfural, a versatile platform chemical made from renewable sources, Chemical Reviews, 113 (2013) 1499-1597
[22] I. van Zandvoort, Y. Wang, C.B. Rasendra, E.R.H. van Eck, P.C.A. Bruijnincx, H.J.Heeres, B.M.Weckhuysen, Formation, molecular structure, and morphology of humins in biomass conversion: influence of feedstock and processing conditions, ChemSusChem, 6 (2013) 1745-1758
[23] S. Handayani, I.S. Arty, Synthesis of Hydroxyl Radical Scavengers from Benzalacetone and its Derivatives, Journal of Physical Science, 19 (2) (2008) 61-68
[24] J.A. Nairn, Materials science& Engineering 5473, Polymer characterization (2003) (http://www.scribd.com; 08-04-2014)
[25] ASTM E 313-05 Standard Practice for Calculating Yellowness and Whiteness Indices from Instrumentally Measured Colour Coordinates: Book of Standards Volume: 06.01 2009.
[26] A. Chrisment, Colour & Colourimetry, Datacolour, Edition 3C (1998)[27] C.P. Maggi, Advantages of Kinematic Viscosity Measurement in Used Oil Analysis, Cannon Company,
A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch
66 67
(http://www.machinerylubrication.com/Read/920/kinematic-viscosity-measurement; 20-06-2012)[28] L.R. Bennedsen, J. Muff, E.G. Sogaard, Influence of chloride and carbonates on the reactivity of activated
persulfate, Chemosphere, 86, 11 (2012) 1092-1097[29] M.Hahn, W. Jaeger, C. Wandrey, G. Reinisch, Zur kinetik der radikalischen polymerization von dimethyl-
diallyl-ammoniumchloride: IV/ Mechanismus von start- und abbruchreaction met persulfat als initiator, Acta Polymerica 35 (5) (1984) 350-358
[30] J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical engineering: Green engineering, Chemical Engineering Journal, 133 (2007) 7-30
[31] ACS Green Chemistry Institute, The twelve principles of green engineering,( www.acs.org; 18-03-2013)[32] ACS Green Chemistry Institute, The twelve principles of green chemistry (www.acs.org; 18-03-2013)[33] Experimental design and Taguchi (http://homepage.ntlworld.com/s.orszulik/index.html; 20-06-2012)
References[1] I. Skeist, Handbook of adhesives, Reinhold Publishing Corp. Chapman & Hall, London, 1962[2] V.A. Lauria, Remoistenable adhesive compositions (1987) US4678824[3] M.S. Mahiel, J.M. Cruden, Adhesive compositions and self-adhesive sheet materials (1988) EP0297900A2[4] M.S. Mahiel, J.M. Cruden, Surface coating compositions (1989) EP0351193A2[5] Jr.J. Wieczorek, L.M. Mahony, Aqueous adhesive compositions for use in binding book (1996) US5519072[6] U. Geissler, H. Hintz, U. Vogt-saggau, Powdery adhesive composition (1997) EP0799876A2[7] T. Mayer, H.P. Weitzel, R. Haerschel, T. Bastelberger, Method for producing polymers stabilized with
protective colloids (2000) US6300403[8] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum
adhesive based on a filled polymer dispersion, WO2006094594A1.[9] K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman,
H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects. To be published.
[10] A.A.M. Maas, J.C.P. Hopman, R.P.W. Kesselmans, Dextrinization of starch (2003) US6613152[11] B. Brimhall, Structure of pyrodextrins, Industrial and engineering chemistry, 36 (1) (1944) 72-75[12] R.J. Dimler, H.A. Davis, G.E. Hilbert, A new anhydride of D-glucose: D-Glucosan ,<1,4> β <1,6>, Journal of
the american chemical society, 68 (7) (1946) 1377-1380[13] R.W. Kerr, F.C. Cleveland, Chemistry of dextrination, Starch, 5 (10) (1953) 261-266[14] M.L. Wolfrom, A. Thompson, R.B. Ward, Composition of pyrodextrins, Industrial and engineering chemistry,
55 (3) (1961) 217-28[15] D.J. Bryce, C.T. Greenwood, Aspects of thermal degradation of starch, Starch, 15 (5) (1963) 166-170[16] A.A. Rosatella, S.P. Simoneonov, R.F.M. Frade, C.A.M. Alfonso, 5-Hydroxymethylfurfural (HMF) as a
building block platform: Biological properties, synthesis and synthetic applications, Green Chemistry, 13 (2011) 754-793
[17] S.K.R. Patil, C.R.F. Lund, Formation and Growth of Humins via Aldol Addition and Condensation during Acid-Catalyzed Conversion of 5-Hydroxymethylfurfural, Energy and Fuels, 25 (2011) 4745-4755
[18] H.E. van Dam, H.P.G. Kieboom,H van Bekkum, The conversion of fructose and glucose in acidic media: Formation of hydroxymethylfurfural, 38 (3) (1986) 95-101
[19] L.A. Ameur, O. Mathieu, V. Lalanne, G. Trystram, I. Birlouez-Aragon, Comparison of the effects of sucrose and hexose on furfural formation and browning in cookies baked at different temperatures. Food Chemistry 101 (2007) 1407–1416.
[20] K. R. Terpstra, A. J. J. Woortman, J. C. P. Hopman, Yellow dextrins: Evaluating changes in structure and colour during processing, Starch, 62 (9) (2010) 449-457
[21] R.J. van Putten, J.C. van der Waal, E. de Jong, C.B. Rasendra, H.J. Heeres, J.G. de Vries, Hydroxymethylfurfural, a versatile platform chemical made from renewable sources, Chemical Reviews, 113 (2013) 1499-1597
[22] I. van Zandvoort, Y. Wang, C.B. Rasendra, E.R.H. van Eck, P.C.A. Bruijnincx, H.J.Heeres, B.M.Weckhuysen, Formation, molecular structure, and morphology of humins in biomass conversion: influence of feedstock and processing conditions, ChemSusChem, 6 (2013) 1745-1758
[23] S. Handayani, I.S. Arty, Synthesis of Hydroxyl Radical Scavengers from Benzalacetone and its Derivatives, Journal of Physical Science, 19 (2) (2008) 61-68
[24] J.A. Nairn, Materials science& Engineering 5473, Polymer characterization (2003) (http://www.scribd.com; 08-04-2014)
[25] ASTM E 313-05 Standard Practice for Calculating Yellowness and Whiteness Indices from Instrumentally Measured Colour Coordinates: Book of Standards Volume: 06.01 2009.
[26] A. Chrisment, Colour & Colourimetry, Datacolour, Edition 3C (1998)[27] C.P. Maggi, Advantages of Kinematic Viscosity Measurement in Used Oil Analysis, Cannon Company,
68
CHAPTER 4Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics
AppendixTable A1: Latex related variables: Viscosity, thermal transitions and DLS based variables .
Code Viscosity Thermal transitions (reversing heat flow) PSD (DLS)
(mPa·s)Tg,onset
(°C)Tg,inflection
(°C)Tg,endset
(°C)DTg
(°C)Diameter
(nm)Half width
(nm)
LLL 40.0 12.3 14.1 16.5 4.2 254 53LLH 36.0 12.0 14.0 16.0 4.1 236 62LHL 34.0 10.8 13.8 15.8 5.0 240 82LHH 32.0 10.7 12.4 14.6 3.9 249 64HLL 40.0 12.0 13.9 15.7 3.8 253 63HLH 36.0 11.3 13.4 15.4 4.1 239 70HHL 34.0 10.8 13.0 14.7 3.9 240 45HHH 32.0 10.4 13.4 15.3 4.9 239 68CCC1 36.0 10.6 12.3 14.9 4.3 238 47CCC2 34.0 10.9 12.8 14.7 3.8 237 53CCC3 34.0 10.6 13.4 15.3 4.7 235 72
Table A2: pVAc composition variables and calculation of f.Code Cold extraction Hot extraction pVAc composition
InsolublepVAc
(mg/g)
SolublepVAc(%)
Precipitatecold
precipitate(%)
pVAc in cold
precipitate(mg/g)
pVAc total
(mg/g)
Homo-polymer(mg/g)
grafted(mg/g)
f(fraction)
Dry matter(mg/g)
LLL 229 202 31,4 72 364 274 91 0.25 431LLH 203 227 31,6 64 365 291 75 0.20 429LHL 273 161 24,5 67 370 228 142 0.38 434LHH 292 133 17,6 51 366 184 182 0.50 425HLL 244 187 24,4 60 366 246 119 0.33 431HLH 229 199 30,8 71 364 270 94 0.26 428HHL 251 177 38,0 95 364 273 92 0.25 428HHH 318 109 21,3 68 368 177 191 0.52 427CCC1 249 181 37,9 94 366 276 90 0.25 431CCC2 225 204 31,9 72 365 276 89 0.24 429CCC3 244 185 28,7 70 365 255 110 0.30 429
68
CHAPTER 4Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics
AppendixTable A1: Latex related variables: Viscosity, thermal transitions and DLS based variables .
Code Viscosity Thermal transitions (reversing heat flow) PSD (DLS)
(mPa·s)Tg,onset
(°C)Tg,inflection
(°C)Tg,endset
(°C)DTg
(°C)Diameter
(nm)Half width
(nm)
LLL 40.0 12.3 14.1 16.5 4.2 254 53LLH 36.0 12.0 14.0 16.0 4.1 236 62LHL 34.0 10.8 13.8 15.8 5.0 240 82LHH 32.0 10.7 12.4 14.6 3.9 249 64HLL 40.0 12.0 13.9 15.7 3.8 253 63HLH 36.0 11.3 13.4 15.4 4.1 239 70HHL 34.0 10.8 13.0 14.7 3.9 240 45HHH 32.0 10.4 13.4 15.3 4.9 239 68CCC1 36.0 10.6 12.3 14.9 4.3 238 47CCC2 34.0 10.9 12.8 14.7 3.8 237 53CCC3 34.0 10.6 13.4 15.3 4.7 235 72
Table A2: pVAc composition variables and calculation of f.Code Cold extraction Hot extraction pVAc composition
InsolublepVAc
(mg/g)
SolublepVAc(%)
Precipitatecold
precipitate(%)
pVAc in cold
precipitate(mg/g)
pVAc total
(mg/g)
Homo-polymer(mg/g)
grafted(mg/g)
f(fraction)
Dry matter(mg/g)
LLL 229 202 31,4 72 364 274 91 0.25 431LLH 203 227 31,6 64 365 291 75 0.20 429LHL 273 161 24,5 67 370 228 142 0.38 434LHH 292 133 17,6 51 366 184 182 0.50 425HLL 244 187 24,4 60 366 246 119 0.33 431HLH 229 199 30,8 71 364 270 94 0.26 428HHL 251 177 38,0 95 364 273 92 0.25 428HHH 318 109 21,3 68 368 177 191 0.52 427CCC1 249 181 37,9 94 366 276 90 0.25 431CCC2 225 204 31,9 72 365 276 89 0.24 429CCC3 244 185 28,7 70 365 255 110 0.30 429
Modified waxy potato starch stabilized polyvinyl acetate latexes
70 71
factually limits the size of the particles (to increase the dry solid content) down to a minimum of 100 nm [8]. High dry solid content latex based on mixing of two monodisperse latexes can also be prepared in one processing step by introducing a second particle nucleation stage during polymerization. This approach requires less processing steps than mixing different latexes and is therefore favoured from a green chemistry and engineering point of view. Secondary particle formation can be initiated by adding emulsifiers after the initial nucleation stage of the polymerization process. The second particle size distribution can be influenced by the time of addition and the particle size decreases with the dosage time. A drawback of this approach is that it increases the amount of material (i.e. emulsifiers) that can migrate out of the polymer layer after application and contaminate the local environment. This migratory risk needs to be minimized according the generally defined rules of producing in a more save and sustainable way. Several indications are already found that more stringent regulations will emerge in the near future [9,10].Pyrodextrins can be used to generate pVAc latexes with high dry solid content up to 70% in a different way [11]. The starch derivative remains in the water phase of the latex and therefore does not increase the volume of the dispersed phase relative to the total volume of the dispersion. The particle size distribution of this type of solid latexes can be selected in a proper range with respect to sensitivity to shear stresses and the resulting viscosity increases. Unfortunately, commonly available latexes are often prepared with emulsifiers, which render them less interesting when making allowances for the migration of these additives during application (vide supra). Moreover, there are also indications that the colouration and chloride content of pyrodextrins affects the polymerization rate to an extent that makes the polymerization process more energy demanding [12]. Maltodextrins might be interesting alternatives because they are less coloured than pyrodextrins and they contain hardly any chloride. However, the exceptional behaviour of pyrodextrins finds its origin in their high degree of degradation (i.e. relatively low molecular weight) and increased density which cannot be matched by commercially available maltodextrins yet [13,14]. Heat during processing can be transferred from the reaction mixture to the reflux cooler by refluxing of the low boiling azeotrope vinyl acetate monomer (VAM) and water (92.7 / 7.3 wt %; 66°C). The addition of VAM to a water based mixture results in a boiling point in the range of 100 - 66°C depending on the composition and the presence of contaminants [14-16]. These considerations are of paramount importance since they clearly indicate that, depending on the chemical composition (in turn related to the conversion of the polymerization reaction) reflux of the liquid might occur resulting in relatively high (and undesirable) energy consumption. On the other hand, this also suggests that the head-space temperature (HST), being directly related to the VAM concentration (and thus the polymerization reaction extent), can be conveniently used to monitor the progress of the reaction in those cases (e.g. industrial settings) in which the use of a sensor is not possible or desirable. In this work we systematically investigated the use of a modified waxy potato starch (maltodextrin) for the stabilization of a pVAc latex in the absence of synthetic emulsifiers [14]. The starch content of the formulation was 25 wt % with respect to pVAc and the total dry solid content was approximately 57 wt %. The selected modified starch has not been investigated before in the literature in the range of composition employed here. The increase in starch
AbstractPolyvinyl acetate latexes with only modified starch as stabilizing agent are interesting alternatives for counterparts based on synthetic additives (i.e. detergents, emulsifiers and/or protective colloids). These latexes do not only have a reduced amount of oil-based products but can be tuned towards relatively high solid contents, thus reducing the amount of water (solvent) needed. This work focused on the synthesis of latexes (57 wt % solid content) stabilized by maltodextrins. In particular it aimed at investigating the impact of changes in temperature (i.e. 75 – 85 °C) and the persulfate amount (i.e. 2 – 4 wt %) during the polymerization on the latex characteristics. The prepared materials displayed viscosities in the range of 700 – 2300 mPa·s and differences in processing conditions were observed as well. Product properties indicated distinct changes in the radical formation process. The viscosity of the latex was related, via a statistical model, to the processing factors mentioned above.
IntroductionStarch containing polyvinyl acetate (pVAc) latexes with dry solid contents higher than 50 wt % (i.e. dry material relative to the total dispersion) are interesting from a commercial and environmental point of view [1-5]. This is not only due to their renewable nature but also with respect to their preparation and application characteristics. Higher dry solid content increases the throughput of the reactor and makes transport of the final product more efficient and cheaper. The presence of less water also diminishes the time for drying and film formation and reduces the energy requirement when drying equipment is required (e.g. in the preparation of plywood).Dry solid contents up to 55 wt % are very common for commercial latexes and modifications in the particle size distribution allow for values up to 75 wt %. However, the viscosity of these latexes is known to increase rapidly once the dry solid content exceeds the 50 wt % level and the latexes become sensitive for shear stress, which might then introduce problems during preparation, handling and/or application [6-8].Latexes with high dry solid content can be designed by calculation of the maximum obtainable value for monodisperse latex. Several models are available for this purpose and the Krieger-Dougherthy equation is shown below [6-8]:
Equation 1:
with being the relative visocisty, the viscosity of the suspension, the viscosity of the medium, the volume fraction of solids in the suspension, the maximum packing fraction and the intrinsic viscosity.Dry solid contents up to 63 wt % are feasible for inert solid spheres arranged in a random close packing (e.g. silica or polyvinyl chloride) [8]. Even higher dry solid contents are accessible by mixing two monodisperse suspensions with a size difference ratio of approximately 4.5. However, the particles in polymer latexes are seldom inert and hard indeformable solids, which limit the applicability of these equations in a number of cases considerably [8]. For example, the effect of inter-particle repulsive forces increases with decreasing particle size and this
Modified waxy potato starch stabilized polyvinyl acetate latexes
70 71
factually limits the size of the particles (to increase the dry solid content) down to a minimum of 100 nm [8]. High dry solid content latex based on mixing of two monodisperse latexes can also be prepared in one processing step by introducing a second particle nucleation stage during polymerization. This approach requires less processing steps than mixing different latexes and is therefore favoured from a green chemistry and engineering point of view. Secondary particle formation can be initiated by adding emulsifiers after the initial nucleation stage of the polymerization process. The second particle size distribution can be influenced by the time of addition and the particle size decreases with the dosage time. A drawback of this approach is that it increases the amount of material (i.e. emulsifiers) that can migrate out of the polymer layer after application and contaminate the local environment. This migratory risk needs to be minimized according the generally defined rules of producing in a more save and sustainable way. Several indications are already found that more stringent regulations will emerge in the near future [9,10].Pyrodextrins can be used to generate pVAc latexes with high dry solid content up to 70% in a different way [11]. The starch derivative remains in the water phase of the latex and therefore does not increase the volume of the dispersed phase relative to the total volume of the dispersion. The particle size distribution of this type of solid latexes can be selected in a proper range with respect to sensitivity to shear stresses and the resulting viscosity increases. Unfortunately, commonly available latexes are often prepared with emulsifiers, which render them less interesting when making allowances for the migration of these additives during application (vide supra). Moreover, there are also indications that the colouration and chloride content of pyrodextrins affects the polymerization rate to an extent that makes the polymerization process more energy demanding [12]. Maltodextrins might be interesting alternatives because they are less coloured than pyrodextrins and they contain hardly any chloride. However, the exceptional behaviour of pyrodextrins finds its origin in their high degree of degradation (i.e. relatively low molecular weight) and increased density which cannot be matched by commercially available maltodextrins yet [13,14]. Heat during processing can be transferred from the reaction mixture to the reflux cooler by refluxing of the low boiling azeotrope vinyl acetate monomer (VAM) and water (92.7 / 7.3 wt %; 66°C). The addition of VAM to a water based mixture results in a boiling point in the range of 100 - 66°C depending on the composition and the presence of contaminants [14-16]. These considerations are of paramount importance since they clearly indicate that, depending on the chemical composition (in turn related to the conversion of the polymerization reaction) reflux of the liquid might occur resulting in relatively high (and undesirable) energy consumption. On the other hand, this also suggests that the head-space temperature (HST), being directly related to the VAM concentration (and thus the polymerization reaction extent), can be conveniently used to monitor the progress of the reaction in those cases (e.g. industrial settings) in which the use of a sensor is not possible or desirable. In this work we systematically investigated the use of a modified waxy potato starch (maltodextrin) for the stabilization of a pVAc latex in the absence of synthetic emulsifiers [14]. The starch content of the formulation was 25 wt % with respect to pVAc and the total dry solid content was approximately 57 wt %. The selected modified starch has not been investigated before in the literature in the range of composition employed here. The increase in starch
AbstractPolyvinyl acetate latexes with only modified starch as stabilizing agent are interesting alternatives for counterparts based on synthetic additives (i.e. detergents, emulsifiers and/or protective colloids). These latexes do not only have a reduced amount of oil-based products but can be tuned towards relatively high solid contents, thus reducing the amount of water (solvent) needed. This work focused on the synthesis of latexes (57 wt % solid content) stabilized by maltodextrins. In particular it aimed at investigating the impact of changes in temperature (i.e. 75 – 85 °C) and the persulfate amount (i.e. 2 – 4 wt %) during the polymerization on the latex characteristics. The prepared materials displayed viscosities in the range of 700 – 2300 mPa·s and differences in processing conditions were observed as well. Product properties indicated distinct changes in the radical formation process. The viscosity of the latex was related, via a statistical model, to the processing factors mentioned above.
IntroductionStarch containing polyvinyl acetate (pVAc) latexes with dry solid contents higher than 50 wt % (i.e. dry material relative to the total dispersion) are interesting from a commercial and environmental point of view [1-5]. This is not only due to their renewable nature but also with respect to their preparation and application characteristics. Higher dry solid content increases the throughput of the reactor and makes transport of the final product more efficient and cheaper. The presence of less water also diminishes the time for drying and film formation and reduces the energy requirement when drying equipment is required (e.g. in the preparation of plywood).Dry solid contents up to 55 wt % are very common for commercial latexes and modifications in the particle size distribution allow for values up to 75 wt %. However, the viscosity of these latexes is known to increase rapidly once the dry solid content exceeds the 50 wt % level and the latexes become sensitive for shear stress, which might then introduce problems during preparation, handling and/or application [6-8].Latexes with high dry solid content can be designed by calculation of the maximum obtainable value for monodisperse latex. Several models are available for this purpose and the Krieger-Dougherthy equation is shown below [6-8]:
Equation 1:
with being the relative visocisty, the viscosity of the suspension, the viscosity of the medium, the volume fraction of solids in the suspension, the maximum packing fraction and the intrinsic viscosity.Dry solid contents up to 63 wt % are feasible for inert solid spheres arranged in a random close packing (e.g. silica or polyvinyl chloride) [8]. Even higher dry solid contents are accessible by mixing two monodisperse suspensions with a size difference ratio of approximately 4.5. However, the particles in polymer latexes are seldom inert and hard indeformable solids, which limit the applicability of these equations in a number of cases considerably [8]. For example, the effect of inter-particle repulsive forces increases with decreasing particle size and this
Modified waxy potato starch stabilized polyvinyl acetate latexes
72 73
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM dosage was also monitored with a balance. A syringe pump was used to add SPS and another syringe pump was used to add STS after the actual polymerization was finished. Differences in dosage amounts were achieved by changing the concentration of solutions used. This approach minimizes the variation in water content during the polymerization and this is important since the initial radical formation takes place in the water phase and is concentration dependent
Statistical designA randomized 22 factorial design with five center points was selected to investigate the influence of two factors (i.e. SPS concentration and the temperature of water bath connected to the reactor (WTR)) on the polymerization reaction [17].
Table 1: Experimental setup and values of selected variablesRun Code Factors
SPS(wt %)
WTR(°C)
1 CC1 3 802 LL1 2 753 CC2 3 804 CC3 3 805 CC4 3 806 HL1 4 757 HH1 4 858 CC5 3 809 LH1 2 85
ProcedureA 25 wt % starch/water mixture was prepared by adding the modified starch to a demi-water containing beaker whilst stirring (3-bladed impeller; 1000 RPM; 10 minutes). The polymerization reactor was filled with 322.8 g of this mixture. Automatic mixing (0-0.75 Hrs: 240 RPM; 0.75-9.0 hrs: 120 RPM) was started and the applied WTR profile and dosage protocols of VAM, SPS and STS are given in Figure 2.VAM was used without inhibitor removal in order to closely match the experimental conditions applicable at industrial level. A total of 0.27 kg VAM was added in all cases and a pre-dosage of 13.5 g VAM was chosen since this ensures a relevant reproducibility level as well as a limited level of reflux [14]. The actual dosage was monitored in time and the amount of VAM added was used for mass balance calculations. The three levels of SPS solution were prepared by weighing 1.00, 1.50 or 2.00 g of SPS in centrifuge tube together with SBC in the ratio 3:4. The weighed SPS and SBC were added to demineralised water and the grand total was 50 g in all cases. The actual addition of the SPS starts after 104 minutes with a pre-dosage of 4.5 ml (4.5 ml/min) followed by 31.5 ml with a dosage speed of 5.25 ml/hr. 2.7 ml 0.3 M STS ml was added with 2.7 ml/hr after the reaction temperature drops below 65 °C during the cooling down. Agitation was continued for at least one hour after the temperature of the water bath
content was expected to have a pronounced effect on product and process characteristics. Information about the impact of changes in these two factors on product and process characteristics is not available in the literature for the selected formulation and this justifies a thorough investigation in this direction. The experiments were designed according to the rules of a 22 factorial augmented with a center point measured in fivefold [17].
ExperimentalMaterialsThe modified waxy potato starch (maltodextrin) used is commercially available under the name of Eliane MD2 (AVEBE U.A.; The Netherlands). The vinyl acetate monomer (VAM) was obtained from ACROS and contains 3-30 ppm hydroquinone as inhibitor. Analytical reagent grade sodium persulfate (SPS) was supplied by VWR International. Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) were both of pro analyse quality and from Merck Germany. STS was added as a 0.3 M solution. 2.0, 3.0 and 4.0 wt % SPS was dissolved in water together with SBC in the ratio 3:4 and. All ingredients were used without additional purification and the solvent was demineralised water in all cases.
EquipmentA double jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles with a width of 1 cm and an outer dimension of 10.5x7 cm (height x diameter)) was applied. The spiral ribbon stirrer was replaced by a rushton impeller (Teflon; 8 cm) in the pH, dissolved oxygen and temperature of the reaction mixture (RMT) experiment. A lid of borosilicate glass with several connection points was placed on top and the reactor was completely insulated with radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the temperature of the headspace in the reactor (HST). The dimensions and settings of the water baths used for temperature control of polymerization reactor and reflux cooler were the same and the temperature difference between inlets and outlets can therefore be used for a heat transport evaluation and energy consumption calculations of reactor and cooler. The feeding lines of VAM and SPS were placed outside the reflux region with the aid of an accessory to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 1).
Figure 1: Schematic representation of the polymerization reactor used.
Modified waxy potato starch stabilized polyvinyl acetate latexes
72 73
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM dosage was also monitored with a balance. A syringe pump was used to add SPS and another syringe pump was used to add STS after the actual polymerization was finished. Differences in dosage amounts were achieved by changing the concentration of solutions used. This approach minimizes the variation in water content during the polymerization and this is important since the initial radical formation takes place in the water phase and is concentration dependent
Statistical designA randomized 22 factorial design with five center points was selected to investigate the influence of two factors (i.e. SPS concentration and the temperature of water bath connected to the reactor (WTR)) on the polymerization reaction [17].
Table 1: Experimental setup and values of selected variablesRun Code Factors
SPS(wt %)
WTR(°C)
1 CC1 3 802 LL1 2 753 CC2 3 804 CC3 3 805 CC4 3 806 HL1 4 757 HH1 4 858 CC5 3 809 LH1 2 85
ProcedureA 25 wt % starch/water mixture was prepared by adding the modified starch to a demi-water containing beaker whilst stirring (3-bladed impeller; 1000 RPM; 10 minutes). The polymerization reactor was filled with 322.8 g of this mixture. Automatic mixing (0-0.75 Hrs: 240 RPM; 0.75-9.0 hrs: 120 RPM) was started and the applied WTR profile and dosage protocols of VAM, SPS and STS are given in Figure 2.VAM was used without inhibitor removal in order to closely match the experimental conditions applicable at industrial level. A total of 0.27 kg VAM was added in all cases and a pre-dosage of 13.5 g VAM was chosen since this ensures a relevant reproducibility level as well as a limited level of reflux [14]. The actual dosage was monitored in time and the amount of VAM added was used for mass balance calculations. The three levels of SPS solution were prepared by weighing 1.00, 1.50 or 2.00 g of SPS in centrifuge tube together with SBC in the ratio 3:4. The weighed SPS and SBC were added to demineralised water and the grand total was 50 g in all cases. The actual addition of the SPS starts after 104 minutes with a pre-dosage of 4.5 ml (4.5 ml/min) followed by 31.5 ml with a dosage speed of 5.25 ml/hr. 2.7 ml 0.3 M STS ml was added with 2.7 ml/hr after the reaction temperature drops below 65 °C during the cooling down. Agitation was continued for at least one hour after the temperature of the water bath
content was expected to have a pronounced effect on product and process characteristics. Information about the impact of changes in these two factors on product and process characteristics is not available in the literature for the selected formulation and this justifies a thorough investigation in this direction. The experiments were designed according to the rules of a 22 factorial augmented with a center point measured in fivefold [17].
ExperimentalMaterialsThe modified waxy potato starch (maltodextrin) used is commercially available under the name of Eliane MD2 (AVEBE U.A.; The Netherlands). The vinyl acetate monomer (VAM) was obtained from ACROS and contains 3-30 ppm hydroquinone as inhibitor. Analytical reagent grade sodium persulfate (SPS) was supplied by VWR International. Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) were both of pro analyse quality and from Merck Germany. STS was added as a 0.3 M solution. 2.0, 3.0 and 4.0 wt % SPS was dissolved in water together with SBC in the ratio 3:4 and. All ingredients were used without additional purification and the solvent was demineralised water in all cases.
EquipmentA double jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles with a width of 1 cm and an outer dimension of 10.5x7 cm (height x diameter)) was applied. The spiral ribbon stirrer was replaced by a rushton impeller (Teflon; 8 cm) in the pH, dissolved oxygen and temperature of the reaction mixture (RMT) experiment. A lid of borosilicate glass with several connection points was placed on top and the reactor was completely insulated with radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the temperature of the headspace in the reactor (HST). The dimensions and settings of the water baths used for temperature control of polymerization reactor and reflux cooler were the same and the temperature difference between inlets and outlets can therefore be used for a heat transport evaluation and energy consumption calculations of reactor and cooler. The feeding lines of VAM and SPS were placed outside the reflux region with the aid of an accessory to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 1).
Figure 1: Schematic representation of the polymerization reactor used.
Modified waxy potato starch stabilized polyvinyl acetate latexes
74 75
Figure 3: The hydrogen ion and oxygen concentration during the first 3 hours of polymerization experiment CC.
The amount of dissolved oxygen increased in the first stage of the dissolution process but gradually returned to 87 % saturation before one hour was passed. The change in oxygen content is probably related to the air trapped in the maltodextrin, which is transferred to the water phase during dissolution process and then slowly removed through the headspace and reflux cooler of the reactor. The pre-dosage of VAM resulted in a sharp decrease in oxygen concentration. This is in line with the assumption that a pre-dosage of VAM is suitable to replace an inert gas purge as pre-polymerization step. The pre-dosage of SPS results in a pronounced increase in oxygen content and shows that it is difficult to execute this type of polymerization in the absence of oxygen. It took approximately one hour to reach a constant pH. The maltodextrin is at this point properly dissolved and the ion exchange between the maltodextrin and water phase is probably in equilibrium as well. After the VAM addition, the pH suddenly dropped. This is most likely caused by the formation of acetic acid originating from the dissociation of VAM.
Reaction scheme 1:
The pre-dosage of the SPS mixture, with a pH of approximately 8.5, results in a sharp increase in pH up to 6.5 (Figure 3). Base catalyzed saponification of VAM is assumed to be acceptable at the observed circumstances. Figure 4 shows HST and water bath power consumption (WPC) consumption profiles of polymerizations performed at a WTR of 75 (LL1) and 85 (HH1) °C. A systematic difference was observed between the HST of LL1 and HH1 just before the VAM dosage was started (1.5 hours) and, at the same point, a systematic difference was observed for the WPC profiles of LL1 and HH1 as well. Both differences are linked to the selected WTR. The HST drops to approximately 65 °C after the dosage of VAM was started and showed a pronounced increase after the dosage was stopped. This phenomenon can be mainly related to the fact that VAM and water forms an azeotrope (VAM / water: 92.7 / 7.3 wt %) with a boiling point of 66 °C, at least if enough VAM is present in the mixture [16]. WPC showed a
reaches a temperature of 20 °C. The dispersion was transferred into a storage container without any additional treatments after this short period of equilibration.
Figure 2: Applied WTR profiles and dosage protocols of SPS/SBC, VAM and STS.
CharacterizationViscosity, pH and dry matter of the final latex were determined with the help of a Brookfield DV-II+( 20RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80 °C) respectively. The hydrogen ion and oxygen content of the reaction mixture during processing were monitored, when applicable, with a P915 controller from LTH equipped with a pH-SterProbe and a DO-electrode 316SS. Ethanal and residual VAM were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 mx0.32 mm) and a flame ionization detector. The gas chromatography measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13000 relative centrifugal force for 10 minutes and the supernatant was mixed 1:1 with 5 mM NaOH. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. Particle size distributions (PSD) were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13-32.5 μm). Fraunhofer theory based calculations were used and the obtained particle size distributions were ISO 13320 compliant. Glass transition temperatures (Tg) were derived from reversing heat flow curves determined with a modulating differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg untreated latex).
ResultsProcess characteristics The hydrogen ion and oxygen concentration of the reaction mixture during the first 3 hours of polymerization CC are given in Figure 3.
Modified waxy potato starch stabilized polyvinyl acetate latexes
74 75
Figure 3: The hydrogen ion and oxygen concentration during the first 3 hours of polymerization experiment CC.
The amount of dissolved oxygen increased in the first stage of the dissolution process but gradually returned to 87 % saturation before one hour was passed. The change in oxygen content is probably related to the air trapped in the maltodextrin, which is transferred to the water phase during dissolution process and then slowly removed through the headspace and reflux cooler of the reactor. The pre-dosage of VAM resulted in a sharp decrease in oxygen concentration. This is in line with the assumption that a pre-dosage of VAM is suitable to replace an inert gas purge as pre-polymerization step. The pre-dosage of SPS results in a pronounced increase in oxygen content and shows that it is difficult to execute this type of polymerization in the absence of oxygen. It took approximately one hour to reach a constant pH. The maltodextrin is at this point properly dissolved and the ion exchange between the maltodextrin and water phase is probably in equilibrium as well. After the VAM addition, the pH suddenly dropped. This is most likely caused by the formation of acetic acid originating from the dissociation of VAM.
Reaction scheme 1:
The pre-dosage of the SPS mixture, with a pH of approximately 8.5, results in a sharp increase in pH up to 6.5 (Figure 3). Base catalyzed saponification of VAM is assumed to be acceptable at the observed circumstances. Figure 4 shows HST and water bath power consumption (WPC) consumption profiles of polymerizations performed at a WTR of 75 (LL1) and 85 (HH1) °C. A systematic difference was observed between the HST of LL1 and HH1 just before the VAM dosage was started (1.5 hours) and, at the same point, a systematic difference was observed for the WPC profiles of LL1 and HH1 as well. Both differences are linked to the selected WTR. The HST drops to approximately 65 °C after the dosage of VAM was started and showed a pronounced increase after the dosage was stopped. This phenomenon can be mainly related to the fact that VAM and water forms an azeotrope (VAM / water: 92.7 / 7.3 wt %) with a boiling point of 66 °C, at least if enough VAM is present in the mixture [16]. WPC showed a
reaches a temperature of 20 °C. The dispersion was transferred into a storage container without any additional treatments after this short period of equilibration.
Figure 2: Applied WTR profiles and dosage protocols of SPS/SBC, VAM and STS.
CharacterizationViscosity, pH and dry matter of the final latex were determined with the help of a Brookfield DV-II+( 20RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80 °C) respectively. The hydrogen ion and oxygen content of the reaction mixture during processing were monitored, when applicable, with a P915 controller from LTH equipped with a pH-SterProbe and a DO-electrode 316SS. Ethanal and residual VAM were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 mx0.32 mm) and a flame ionization detector. The gas chromatography measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13000 relative centrifugal force for 10 minutes and the supernatant was mixed 1:1 with 5 mM NaOH. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. Particle size distributions (PSD) were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13-32.5 μm). Fraunhofer theory based calculations were used and the obtained particle size distributions were ISO 13320 compliant. Glass transition temperatures (Tg) were derived from reversing heat flow curves determined with a modulating differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg untreated latex).
ResultsProcess characteristics The hydrogen ion and oxygen concentration of the reaction mixture during the first 3 hours of polymerization CC are given in Figure 3.
Modified waxy potato starch stabilized polyvinyl acetate latexes
76 77
Figure 6: Profiles DTR and DTC for experiments HH1 (85°C) and LL1 (75°C).
This boiling did not only transfer heat from the reaction mixture to the cooler but can also induce significant changes in distribution and concentration of the reaction mixture [18]. A small change in DTC for HH1 was observed after approximately 4.5 hours, which might be linked to a reduction in free VAM of the water phase due to its migration to the pVAc particles. This is in agreement with the fact that polymer particles can, when they reach a certain size, store significant amounts of monomer [19-21]. A logical consequence of this observation is that the size of the particle is related to the amount of free monomer that migrates into the polymeric phase. Therefore, changes in free monomer content can significantly affect the viscosity level of the reaction mixture, which constitutes an undesired effect if such thickening effect results in a poor level of mixing. The level of torque (proportional to the viscosity of the reaction mixture) was around 55 Ncm at the initial stage of the polymerization and showed a small gradual increase during polymerization (Figure 7). The torque profiles of all latexes fall within a range equal to twice the σ of the CC settings, thus indicating that no extreme changes in viscosities took place during processing.
Figure 7: RPM and corresponding torque (average of the CC experiments) as function of the reaction time with actual RPM and level of torque at initial stage of polymerization. Error bars represent two times σ of the average torque of the CC experiments.
small increase during the dosage of VAM. This can probably be linked to an increased level of refluxing. This is also in line with the assumption that the RMT is close to 66 °C during the VAM dosage. This close relationship between HST and RMT is shown in Figure 5. The manual addition of 35ml VAM was carried out in order to show the impact of a sudden change in VAM and water on HST and RMT.
Figure 4: HST during processing for experiment HH1 (85°C) and LL1 (75°) together with the corresponding WPC profiles.
Figure 5: Correlation between HST and RMT as function of the time. The dotted line represents the VAM dosage profile during the polymerization experiment CC.
This confirms that the two temperatures (i.e. RMT and HST) are intimately related and that the latter can in principle be used to monitor the progress of the reaction in those cases in which a direct use of a sensor is practically unfeasible. The temperature difference between inlet and outlet of the reactor and cooler are defined as DTR and DTC respectively. DTR and DTC of experiments HH1 and LL1 appeared to be similar (~0.7 °C) up to 4.5 hrs and suggests that reflux based heat transfer was the most important factor in heat loss process (Figure 6).
Modified waxy potato starch stabilized polyvinyl acetate latexes
76 77
Figure 6: Profiles DTR and DTC for experiments HH1 (85°C) and LL1 (75°C).
This boiling did not only transfer heat from the reaction mixture to the cooler but can also induce significant changes in distribution and concentration of the reaction mixture [18]. A small change in DTC for HH1 was observed after approximately 4.5 hours, which might be linked to a reduction in free VAM of the water phase due to its migration to the pVAc particles. This is in agreement with the fact that polymer particles can, when they reach a certain size, store significant amounts of monomer [19-21]. A logical consequence of this observation is that the size of the particle is related to the amount of free monomer that migrates into the polymeric phase. Therefore, changes in free monomer content can significantly affect the viscosity level of the reaction mixture, which constitutes an undesired effect if such thickening effect results in a poor level of mixing. The level of torque (proportional to the viscosity of the reaction mixture) was around 55 Ncm at the initial stage of the polymerization and showed a small gradual increase during polymerization (Figure 7). The torque profiles of all latexes fall within a range equal to twice the σ of the CC settings, thus indicating that no extreme changes in viscosities took place during processing.
Figure 7: RPM and corresponding torque (average of the CC experiments) as function of the reaction time with actual RPM and level of torque at initial stage of polymerization. Error bars represent two times σ of the average torque of the CC experiments.
small increase during the dosage of VAM. This can probably be linked to an increased level of refluxing. This is also in line with the assumption that the RMT is close to 66 °C during the VAM dosage. This close relationship between HST and RMT is shown in Figure 5. The manual addition of 35ml VAM was carried out in order to show the impact of a sudden change in VAM and water on HST and RMT.
Figure 4: HST during processing for experiment HH1 (85°C) and LL1 (75°) together with the corresponding WPC profiles.
Figure 5: Correlation between HST and RMT as function of the time. The dotted line represents the VAM dosage profile during the polymerization experiment CC.
This confirms that the two temperatures (i.e. RMT and HST) are intimately related and that the latter can in principle be used to monitor the progress of the reaction in those cases in which a direct use of a sensor is practically unfeasible. The temperature difference between inlet and outlet of the reactor and cooler are defined as DTR and DTC respectively. DTR and DTC of experiments HH1 and LL1 appeared to be similar (~0.7 °C) up to 4.5 hrs and suggests that reflux based heat transfer was the most important factor in heat loss process (Figure 6).
Modified waxy potato starch stabilized polyvinyl acetate latexes
78 79
evaporates therefore easily during preparation, storage or sample treatment. The value found for ethanal might therefore be considerably lower than the actual amount formed during processing. The observed increase in acetate concentration with initiator intake is plausible due to the fact that the ratio between amounts of SPS and SBC is fixed. An increase in SPS was therefore coupled with an increase in the amount of added SBC. This influences the reaction pH and, as a consequence, the VAM hydrolysis. Part of the acetate might also originate from the hydrolysis of pVAc, which remained in the water phase after the centrifugation step. Thiosulfate was added in the final stage of the polymerization in order to convert remnant SPS. The stoichiometry of the redox equation of thiosulfate and SPS is 2 to 1 [22].
Reaction scheme 1:
About 0.02 mmol thiosulfate, of the 0.81 mmol added, was still present in the latexes and this justified the assumption that all SPS was converted. All latexes had similar thiosulfate concentrations and this is indicative for a similar SPS concentration at the point of thiosulfate dosage (10 hrs after initiation of the experiment). The sulfate concentration was influenced by both investigated factors and their interaction. The observed differences probably originate from a combination of amounts of SPS were added and changes in radical formation process. The thermal dissociation of SPS was not only influenced by temperature but by changes in composition of the reaction mixture as well [23-29]. However, the experimental setup does not allow solid conclusions in this area and additional research is required to address this aspect of the polymerization properly. The volume mean diameter (VMD), thermal transitions (Tg,onset,Tg,inflection and Tg,endset) and viscosity of the latexes were also determined (Table 4). Latex prepared with the setting HL1 and its counterpart HH1 appeared to differ considerably from the average PSD but coincide with the PSD’s CC1 and CC4 respectively. The observed differences in PSD can therefore be considered as noise of the experiment. No significant differences were observed between the determined Tg’s and this was indicative for a similar composition and structure of the pVAc formed in all cases. The level of pVAc branching (theoretically a crucial factor influencing the Tg values) is known to be proportional with RMT and the assumption that all polymerization are executed at similar RMT’s is therefore plausible [20,30].
Table 4: Product characteristics VMD, thermal transitions and viscosity.Run Code Latex properties Thermal transitions (Reversing heat flow)
VMD(mm)
Viscosity(mPa·s)
Tg,onset (°C)
Tg,inflection(°C)
Tg,endset (°C)
DTg(°C)
1 CC1 1.7 1585 7.2 9.3 12.8 5.62 LL1 1.8 680 9.9 12.7 14.8 4.93 CC2 1.9 1485 8.8 10.8 13.1 4.34 CC3 1.9 1460 8.3 10.2 12.9 4.65 CC4 1.5 1760 8.9 11.2 13.7 4.86 HL1 1.9 1170 8.8 10.8 13.2 4.47 HH1 1.5 1600 7.0 9.3 12.1 5.18 CC5 1.8 1685 8.0 10.1 12.6 4.69 LH1 1.7 2330 8.4 10.6 12.5 4.1
Latex composition and propertiesThe latex characteristics and chemical composition (including dry matter, ethanal, acetate, thiosulfate and sulfate contents) were determined and evaluated using Pareto calculations (Table 2 and 3). All dry matter recoveries exceeded 97% and the small differences cannot be deduced to the applied reaction conditions according to the calculations performed. Some changes in final pH were significantly linked to the factors investigated but the observed variation was only modest and did not exceed the region 4.8 - 5.0. An increase in SPS intake leads to a decrease in residual VAM (negative sign of the corresponding effect in Table 3). This can be explained by the fact that more radicals during processing lead to less unreacted VAM in the final latex. On the other hand, the amount of residual VAM was positively influenced by WTR. This is not expected since SPS is used in thermal dissociation mode and the radical generation during processing increases with the RMT. One might then speculate that WTR is therefore not linked to the RMT, which is in corroborated by the observation that the interaction between SPS and WTR is not statistically significant. The observed behaviour is in line with the hypothesis that all polymerizations are executed at the boiling temperature of the azeotrope water and VAM mainly (vide supra).
Table 2: Product characteristics: Dry matter, pH, ethanal, VAM, acetate, sulfate and thiosulfate content.Run Code Latex Chemical composition
Dry matter Recovered
(%)pH Ethanal
(mg/g)VAM
(mg/g)Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol)
1 CC1 98.8 4.9 2.58 7.26 4.88 0.25 0.0212 LL1 99.1 4.9 1.87 6.35 4.88 0.25 0.0213 CC2 97.7 4.9 1.82 6.77 4.85 0.25 0.0214 CC3 98.1 4.9 2.13 7.23 5.05 0.26 0.0255 CC4 97.4 4.9 2.06 7.71 4.74 0.26 0.0206 HL1 97.7 5.0 3.02 5.89 5.39 0.35 0.0217 HH1 97.8 4.9 3.00 8.62 5.48 0.34 0.0238 CC5 98.3 4.9 2.75 7.06 4.71 0.26 0.0249 LH1 98.4 4.8 2.34 10.54 4.57 0.17 0.023
Table 3.Pareto calculations: Dry matter, pH, ethanal, VAM, acetate, sulfate and thiosulfate content (x = not significant).
Factors Dry matter recovered pH Ethanal VAM Acetate Sulfate Thiosulfate
SPS x 0.1 x -0.12 0.71 0.14 xWTR x -0.1 x 0.35 x -0.48 xSPS*WTR x x x x x 0.04 x
Residual ethanal and acetate were not influenced in the same way by the changes in SPS and WTR. This cannot be explained only by the hydrolysis of VAM, which would result in equal amounts of acetate and ethanal (vide supra). The boiling point of ethanal is 21°C and
Modified waxy potato starch stabilized polyvinyl acetate latexes
78 79
evaporates therefore easily during preparation, storage or sample treatment. The value found for ethanal might therefore be considerably lower than the actual amount formed during processing. The observed increase in acetate concentration with initiator intake is plausible due to the fact that the ratio between amounts of SPS and SBC is fixed. An increase in SPS was therefore coupled with an increase in the amount of added SBC. This influences the reaction pH and, as a consequence, the VAM hydrolysis. Part of the acetate might also originate from the hydrolysis of pVAc, which remained in the water phase after the centrifugation step. Thiosulfate was added in the final stage of the polymerization in order to convert remnant SPS. The stoichiometry of the redox equation of thiosulfate and SPS is 2 to 1 [22].
Reaction scheme 1:
About 0.02 mmol thiosulfate, of the 0.81 mmol added, was still present in the latexes and this justified the assumption that all SPS was converted. All latexes had similar thiosulfate concentrations and this is indicative for a similar SPS concentration at the point of thiosulfate dosage (10 hrs after initiation of the experiment). The sulfate concentration was influenced by both investigated factors and their interaction. The observed differences probably originate from a combination of amounts of SPS were added and changes in radical formation process. The thermal dissociation of SPS was not only influenced by temperature but by changes in composition of the reaction mixture as well [23-29]. However, the experimental setup does not allow solid conclusions in this area and additional research is required to address this aspect of the polymerization properly. The volume mean diameter (VMD), thermal transitions (Tg,onset,Tg,inflection and Tg,endset) and viscosity of the latexes were also determined (Table 4). Latex prepared with the setting HL1 and its counterpart HH1 appeared to differ considerably from the average PSD but coincide with the PSD’s CC1 and CC4 respectively. The observed differences in PSD can therefore be considered as noise of the experiment. No significant differences were observed between the determined Tg’s and this was indicative for a similar composition and structure of the pVAc formed in all cases. The level of pVAc branching (theoretically a crucial factor influencing the Tg values) is known to be proportional with RMT and the assumption that all polymerization are executed at similar RMT’s is therefore plausible [20,30].
Table 4: Product characteristics VMD, thermal transitions and viscosity.Run Code Latex properties Thermal transitions (Reversing heat flow)
VMD(mm)
Viscosity(mPa·s)
Tg,onset (°C)
Tg,inflection(°C)
Tg,endset (°C)
DTg(°C)
1 CC1 1.7 1585 7.2 9.3 12.8 5.62 LL1 1.8 680 9.9 12.7 14.8 4.93 CC2 1.9 1485 8.8 10.8 13.1 4.34 CC3 1.9 1460 8.3 10.2 12.9 4.65 CC4 1.5 1760 8.9 11.2 13.7 4.86 HL1 1.9 1170 8.8 10.8 13.2 4.47 HH1 1.5 1600 7.0 9.3 12.1 5.18 CC5 1.8 1685 8.0 10.1 12.6 4.69 LH1 1.7 2330 8.4 10.6 12.5 4.1
Latex composition and propertiesThe latex characteristics and chemical composition (including dry matter, ethanal, acetate, thiosulfate and sulfate contents) were determined and evaluated using Pareto calculations (Table 2 and 3). All dry matter recoveries exceeded 97% and the small differences cannot be deduced to the applied reaction conditions according to the calculations performed. Some changes in final pH were significantly linked to the factors investigated but the observed variation was only modest and did not exceed the region 4.8 - 5.0. An increase in SPS intake leads to a decrease in residual VAM (negative sign of the corresponding effect in Table 3). This can be explained by the fact that more radicals during processing lead to less unreacted VAM in the final latex. On the other hand, the amount of residual VAM was positively influenced by WTR. This is not expected since SPS is used in thermal dissociation mode and the radical generation during processing increases with the RMT. One might then speculate that WTR is therefore not linked to the RMT, which is in corroborated by the observation that the interaction between SPS and WTR is not statistically significant. The observed behaviour is in line with the hypothesis that all polymerizations are executed at the boiling temperature of the azeotrope water and VAM mainly (vide supra).
Table 2: Product characteristics: Dry matter, pH, ethanal, VAM, acetate, sulfate and thiosulfate content.Run Code Latex Chemical composition
Dry matter Recovered
(%)pH Ethanal
(mg/g)VAM
(mg/g)Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol)
1 CC1 98.8 4.9 2.58 7.26 4.88 0.25 0.0212 LL1 99.1 4.9 1.87 6.35 4.88 0.25 0.0213 CC2 97.7 4.9 1.82 6.77 4.85 0.25 0.0214 CC3 98.1 4.9 2.13 7.23 5.05 0.26 0.0255 CC4 97.4 4.9 2.06 7.71 4.74 0.26 0.0206 HL1 97.7 5.0 3.02 5.89 5.39 0.35 0.0217 HH1 97.8 4.9 3.00 8.62 5.48 0.34 0.0238 CC5 98.3 4.9 2.75 7.06 4.71 0.26 0.0249 LH1 98.4 4.8 2.34 10.54 4.57 0.17 0.023
Table 3.Pareto calculations: Dry matter, pH, ethanal, VAM, acetate, sulfate and thiosulfate content (x = not significant).
Factors Dry matter recovered pH Ethanal VAM Acetate Sulfate Thiosulfate
SPS x 0.1 x -0.12 0.71 0.14 xWTR x -0.1 x 0.35 x -0.48 xSPS*WTR x x x x x 0.04 x
Residual ethanal and acetate were not influenced in the same way by the changes in SPS and WTR. This cannot be explained only by the hydrolysis of VAM, which would result in equal amounts of acetate and ethanal (vide supra). The boiling point of ethanal is 21°C and
Modified waxy potato starch stabilized polyvinyl acetate latexes
80 81
2 and 3). The viscosity of the latex was correlated with its residual VAM content because viscosity was strongly related with the particle size, which in turn increased with the residual VAM content. Increases in particle size up to a factor of 7 are reported [31]. This is nicely in agreement (vide supra) with the hypothesis of monomer migration from the water to the polymeric phase (particles) as deduced from the reflux profile during the polymerization process. Moreover, there were also indications that the latex viscosity was linked to the residual VAM content. However, this correlation was poor and the presence of outliers cannot be excluded yet (Figure 9). More detailed investigations on the mechanism (at molecular level) are needed to further clarify this point; it is worth noticing how both process and product characteristics seem in this case to consistently indicate such a mechanism.
Figure 9: Latex viscosity as function of its residual VAM content.
ConclusionsThe viscosity of waxy potato maltodextrin protected polyvinyl acetate latex (57 wt % dry matter) can be influenced by changes in water bath temperature and initiator concentration during processing. The applied reaction conditions resulted in latex viscosities ranging from 700 to 2300 mPa·s. The maltodextrin used is investigated before but not in this particular ratio modified starch to polyvinyl acetate (i.e. 25 wt % on polyvinyl acetate). The differences in latex compositions were indicative for changes in the radical formation process while both process and product characteristics indicate the presence of vinyl acetate monomer in the polymeric particles (during the polymerization). The processing conditions also resulted in differences in vinyl acetate monomer content and this variable has an influence on the latex viscosity. Observed variations in viscosity can be deduced to either differences in residual vinyl acetate monomer, degree of maltodextrin degradation during processing or combinations thereof. Additional research is required if the actual contribution of each parameter needs to be known exactly. On the other hand, the observed data demonstrates the feasibility of the original idea (i.e. use of maltodextrin as protective colloid in relatively high solid-content latexes) and also the robustness of the applied polymerization settings in terms of process (e.g. viscosity) and product (e.g. thermal) properties.
The VMD of the latexes were not significantly affected by the applied settings according to the Pareto calculations and the corresponding PSD’s did not show significant differences either (Figure 8).
Figure 8: PSD’s of the experiments HH1, HL1 and the average of the CC experiment (error bars representing 2 times σ of the CC experiment).
The impact of WTR on the latex viscosity (as represented by the sign and magnitude of the corresponding statistical coefficients, not shown for brevity) was positive on its own and negative in combination with SPS. The effect of WTR on the actual RMT was probably small, or maybe even absent, due to reflux based heat transport from the RMT to the cooler and can therefore not be a direct cause for the proportional relationship between WTR and viscosity (section 3.1 & 3.2). WTR, and its interaction with SPS, had no influence on the measured PSD of the latex also and the observed changes in viscosity might therefore originated from alterations in the water phase composition. Indeed, a change in degree of degradation of the starch derivative present in the water phase might be a proper explanation for the observed behaviour even if additional research is needed to validate this assumption properly.Factorial calculations (Minitab 16) allow quantification of the effect of process conditions on latex viscosity (Equation 2). A difference of 4.2% between R2 (95.9 %) and R2
adj (91.7 %) was observed, thus indicating that the corresponding model is acceptable even if it does not fit all information totally.
Equation 2:
Latex viscosity = - 21335 + 4820*[SPS] + 287*[WTR] - 61*[SPS] *[WTR] + 150.
The observed deviation between R2 and R2adj might originate from the absence of a relevant
factor in the statistical analysis, namely the amount of residual VAM. This response was significantly influenced by the applied changes in WTR and initiator concentration (Table
Modified waxy potato starch stabilized polyvinyl acetate latexes
80 81
2 and 3). The viscosity of the latex was correlated with its residual VAM content because viscosity was strongly related with the particle size, which in turn increased with the residual VAM content. Increases in particle size up to a factor of 7 are reported [31]. This is nicely in agreement (vide supra) with the hypothesis of monomer migration from the water to the polymeric phase (particles) as deduced from the reflux profile during the polymerization process. Moreover, there were also indications that the latex viscosity was linked to the residual VAM content. However, this correlation was poor and the presence of outliers cannot be excluded yet (Figure 9). More detailed investigations on the mechanism (at molecular level) are needed to further clarify this point; it is worth noticing how both process and product characteristics seem in this case to consistently indicate such a mechanism.
Figure 9: Latex viscosity as function of its residual VAM content.
ConclusionsThe viscosity of waxy potato maltodextrin protected polyvinyl acetate latex (57 wt % dry matter) can be influenced by changes in water bath temperature and initiator concentration during processing. The applied reaction conditions resulted in latex viscosities ranging from 700 to 2300 mPa·s. The maltodextrin used is investigated before but not in this particular ratio modified starch to polyvinyl acetate (i.e. 25 wt % on polyvinyl acetate). The differences in latex compositions were indicative for changes in the radical formation process while both process and product characteristics indicate the presence of vinyl acetate monomer in the polymeric particles (during the polymerization). The processing conditions also resulted in differences in vinyl acetate monomer content and this variable has an influence on the latex viscosity. Observed variations in viscosity can be deduced to either differences in residual vinyl acetate monomer, degree of maltodextrin degradation during processing or combinations thereof. Additional research is required if the actual contribution of each parameter needs to be known exactly. On the other hand, the observed data demonstrates the feasibility of the original idea (i.e. use of maltodextrin as protective colloid in relatively high solid-content latexes) and also the robustness of the applied polymerization settings in terms of process (e.g. viscosity) and product (e.g. thermal) properties.
The VMD of the latexes were not significantly affected by the applied settings according to the Pareto calculations and the corresponding PSD’s did not show significant differences either (Figure 8).
Figure 8: PSD’s of the experiments HH1, HL1 and the average of the CC experiment (error bars representing 2 times σ of the CC experiment).
The impact of WTR on the latex viscosity (as represented by the sign and magnitude of the corresponding statistical coefficients, not shown for brevity) was positive on its own and negative in combination with SPS. The effect of WTR on the actual RMT was probably small, or maybe even absent, due to reflux based heat transport from the RMT to the cooler and can therefore not be a direct cause for the proportional relationship between WTR and viscosity (section 3.1 & 3.2). WTR, and its interaction with SPS, had no influence on the measured PSD of the latex also and the observed changes in viscosity might therefore originated from alterations in the water phase composition. Indeed, a change in degree of degradation of the starch derivative present in the water phase might be a proper explanation for the observed behaviour even if additional research is needed to validate this assumption properly.Factorial calculations (Minitab 16) allow quantification of the effect of process conditions on latex viscosity (Equation 2). A difference of 4.2% between R2 (95.9 %) and R2
adj (91.7 %) was observed, thus indicating that the corresponding model is acceptable even if it does not fit all information totally.
Equation 2:
Latex viscosity = - 21335 + 4820*[SPS] + 287*[WTR] - 61*[SPS] *[WTR] + 150.
The observed deviation between R2 and R2adj might originate from the absence of a relevant
factor in the statistical analysis, namely the amount of residual VAM. This response was significantly influenced by the applied changes in WTR and initiator concentration (Table
Modified waxy potato starch stabilized polyvinyl acetate latexes
82 83
References[1]. ACS Green Chemistry Institute, The twelve principles of green engineering,( www.acs.org; 18-03-2013)[2]. ACS Green Chemistry Institute, The twelve principles of green chemistry (www.acs.org;18-03-2013)[3]. J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical
engineering: Green engineering, Chemical Engineering Journal, 133 ( 2007) 7-30[4]. M. Demendonca, T.E. Baxter, Design for the environment (DFE), An approach to achieve the ISO-14000
international standardization, Environmental management and health, 12 (1) (2001) 51-56[5]. P.T. Anastas, J.C. Warner, Green Chemistry: Theory and Practice, Oxford University Press: New York, 1998[6]. A. Guyot, F. Chu, M. Schneider, C. Graillat, T.F. McKenna, High solid content latexes, Prog Polym Sci, 27
(2002) 1573-1615[7]. F. Chu, A. Guyot, High solids content latexes with low viscosity, Colloid Polym Sci, 279 (2001) 361 [8]. P. Rolfe, J. Langridge, An Overview of the Use of Rheology for Adhesive Manufacturers, Malvern Instruments
( http://www.pstc.org/files/public/Rolfe.pdf; 20-06-2012)[9]. Henkel, Water-based Adhesin adhesives as low migration products, ( http://www.henkel.com/com/content_
data/20120503_Drupa_Adhesin_Low_Migration_eng.pdf;7-12-2012)[10]. P. Jones, Dispersion polymers: Technology & applications, 22 (11) 2011, 1[11]. O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer,Gum
adhesive based on a filled polymer dispersion, WO2006094594A1.[12]. K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, A systematic study on synthesis and
properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch, To be published[13]. K. R. Terpstra, A. J. J. Woortman, J. C. P. Hopman, Yellow dextrins: Evaluating changes in structure and
colour during processing,Starch, 62 (9) (2010) 449-457[14]. K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman,
H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects, To be published
[15]. E.K. Hilmen, Separation of azeotropc mixtures: Tools for analysis and studies on batch distillation operation (http://www.diva-portal.org/smash/get/diva2:125421/FULLTEXT01.pdf; 08-04-2014))
[16]. Vinyl acetate council, (http://www.vinylacetate.org/properties.pdf; 20-06-2012).[17]. D.C. Montgomery, Design and analysis of experiments, 8th edition, John Wiley & Sons Inc., Hoboken, 2013[18]. R.L. Zollars, C.T. Chen, D.A. Jones, Distribution of volatile species in a refluxing polymer colloid, AIChE. J.,
34 (5) (1988) 733-742.[19]. J. Ugelstad, F.K. Hansen, Kinetics and mechanism of emulsion polymerization, Rubber Chem. Techn, 49
(1976) 536-609.[20]. H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/
bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).[21]. W.J. Priest, Particle growth in the aqueous polymerization of vinyl acetate, J. Phys. Chem, 56 (1952) 1077[22]. J.P. Riggs, F. Rodriguez, Polymerization of acrylamide initiated by the persulfate-thiosulfate redox couple,
J. Polym. Sci. Chem Ed. 5 (1967) 3167-3181.[23]. A.M. Santos, P.H. Vindevoghel, C. Graillat, A. Guyot, J. Guillot ,Study of thermal decomposition of potassium
persulfate by potentiometry and capillary electrophoresis, J Appl Polym Sci, Part A Polym Chem, 34 (1996) 1271
[24]. K.Y. van Berkel, G.T. Russel, R.G. Gilbert, The dissociation rate coefficient of persulfate in emulsion polymerization systems, Polymer, 47 (2006) 4667
[25]. B.W. Brooks, B.O. Makanjuola, The rate of persulfate decomposition in the presence of polymer latices Makromol Chem Rapid Commun, 2 (1981) 69
[26]. H. de Bruyn, R.G. Gilbert, Induced decomposition of persulfate by vinyl acetate, 42 (2001) 7999[27]. D.A. House, Kinetics and mechanism of oxidations by peroxydisulfate, Chem Rev, 62 (1962) 185
AbbreviationspVAc : Polyvinyl acetate.VAM : Vinyl acetate monomer.HST : Headspace temperature.SPS : Sodium persulfate.SBC : Sodium bicarbonate.STS : Sodium thiosulfate.WTR : Water bath temperature reactor.L : Low level.H : High level.C : Centre level. RPM : Revolutions per minute.PSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulating differential scanning calorimeter.WPC : Water bath power consumption.RMT : Reaction mixture temperature.DTR : Temperature “inlet reactor “– “outlet reactor”. DTC : Temperature “inlet cooler” – “outlet cooler”.σ : Standard deviation.VMD : Volume Mean Diameter.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
Modified waxy potato starch stabilized polyvinyl acetate latexes
82 83
References[1]. ACS Green Chemistry Institute, The twelve principles of green engineering,( www.acs.org; 18-03-2013)[2]. ACS Green Chemistry Institute, The twelve principles of green chemistry (www.acs.org;18-03-2013)[3]. J. Garcia-Serna, L. Perez-Barrigon, M.J. Cocero, New trends for design towards sustainability in chemical
engineering: Green engineering, Chemical Engineering Journal, 133 ( 2007) 7-30[4]. M. Demendonca, T.E. Baxter, Design for the environment (DFE), An approach to achieve the ISO-14000
international standardization, Environmental management and health, 12 (1) (2001) 51-56[5]. P.T. Anastas, J.C. Warner, Green Chemistry: Theory and Practice, Oxford University Press: New York, 1998[6]. A. Guyot, F. Chu, M. Schneider, C. Graillat, T.F. McKenna, High solid content latexes, Prog Polym Sci, 27
(2002) 1573-1615[7]. F. Chu, A. Guyot, High solids content latexes with low viscosity, Colloid Polym Sci, 279 (2001) 361 [8]. P. Rolfe, J. Langridge, An Overview of the Use of Rheology for Adhesive Manufacturers, Malvern Instruments
( http://www.pstc.org/files/public/Rolfe.pdf; 20-06-2012)[9]. Henkel, Water-based Adhesin adhesives as low migration products, ( http://www.henkel.com/com/content_
data/20120503_Drupa_Adhesin_Low_Migration_eng.pdf;7-12-2012)[10]. P. Jones, Dispersion polymers: Technology & applications, 22 (11) 2011, 1[11]. O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer,Gum
adhesive based on a filled polymer dispersion, WO2006094594A1.[12]. K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, A systematic study on synthesis and
properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch, To be published[13]. K. R. Terpstra, A. J. J. Woortman, J. C. P. Hopman, Yellow dextrins: Evaluating changes in structure and
colour during processing,Starch, 62 (9) (2010) 449-457[14]. K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman,
H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects, To be published
[15]. E.K. Hilmen, Separation of azeotropc mixtures: Tools for analysis and studies on batch distillation operation (http://www.diva-portal.org/smash/get/diva2:125421/FULLTEXT01.pdf; 08-04-2014))
[16]. Vinyl acetate council, (http://www.vinylacetate.org/properties.pdf; 20-06-2012).[17]. D.C. Montgomery, Design and analysis of experiments, 8th edition, John Wiley & Sons Inc., Hoboken, 2013[18]. R.L. Zollars, C.T. Chen, D.A. Jones, Distribution of volatile species in a refluxing polymer colloid, AIChE. J.,
34 (5) (1988) 733-742.[19]. J. Ugelstad, F.K. Hansen, Kinetics and mechanism of emulsion polymerization, Rubber Chem. Techn, 49
(1976) 536-609.[20]. H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/
bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).[21]. W.J. Priest, Particle growth in the aqueous polymerization of vinyl acetate, J. Phys. Chem, 56 (1952) 1077[22]. J.P. Riggs, F. Rodriguez, Polymerization of acrylamide initiated by the persulfate-thiosulfate redox couple,
J. Polym. Sci. Chem Ed. 5 (1967) 3167-3181.[23]. A.M. Santos, P.H. Vindevoghel, C. Graillat, A. Guyot, J. Guillot ,Study of thermal decomposition of potassium
persulfate by potentiometry and capillary electrophoresis, J Appl Polym Sci, Part A Polym Chem, 34 (1996) 1271
[24]. K.Y. van Berkel, G.T. Russel, R.G. Gilbert, The dissociation rate coefficient of persulfate in emulsion polymerization systems, Polymer, 47 (2006) 4667
[25]. B.W. Brooks, B.O. Makanjuola, The rate of persulfate decomposition in the presence of polymer latices Makromol Chem Rapid Commun, 2 (1981) 69
[26]. H. de Bruyn, R.G. Gilbert, Induced decomposition of persulfate by vinyl acetate, 42 (2001) 7999[27]. D.A. House, Kinetics and mechanism of oxidations by peroxydisulfate, Chem Rev, 62 (1962) 185
AbbreviationspVAc : Polyvinyl acetate.VAM : Vinyl acetate monomer.HST : Headspace temperature.SPS : Sodium persulfate.SBC : Sodium bicarbonate.STS : Sodium thiosulfate.WTR : Water bath temperature reactor.L : Low level.H : High level.C : Centre level. RPM : Revolutions per minute.PSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulating differential scanning calorimeter.WPC : Water bath power consumption.RMT : Reaction mixture temperature.DTR : Temperature “inlet reactor “– “outlet reactor”. DTC : Temperature “inlet cooler” – “outlet cooler”.σ : Standard deviation.VMD : Volume Mean Diameter.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
84
CHAPTER 5Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and waxy potato starch
[28]. C. Liang, H.W. Su, Identification of sulfate and hydroxyl radicals in thermally activated persulfate, Ind Eng Chem Res, 48 (2009) 5558
[29]. S.M. Sankalia, D.K. Chaudhuri, J.J. Hermans, Studies on the mechanism of persulfate initiated grafting on cellulose, Can J of Chem, 1962, 40: 2249
[30]. D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837.
[31]. H. B. Yamak, Emulsion Polymerization: Effects of Polymerization Variables on the Properties of Vinyl Acetate Based Emulsion Polymers, Polymer Science, Dr. Faris Yılmaz (Ed.), Intech, 2013 (http://www.intechopen.com/books/polymer-science; 20-12-2013)
84
CHAPTER 5Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and waxy potato starch
[28]. C. Liang, H.W. Su, Identification of sulfate and hydroxyl radicals in thermally activated persulfate, Ind Eng Chem Res, 48 (2009) 5558
[29]. S.M. Sankalia, D.K. Chaudhuri, J.J. Hermans, Studies on the mechanism of persulfate initiated grafting on cellulose, Can J of Chem, 1962, 40: 2249
[30]. D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837.
[31]. H. B. Yamak, Emulsion Polymerization: Effects of Polymerization Variables on the Properties of Vinyl Acetate Based Emulsion Polymers, Polymer Science, Dr. Faris Yılmaz (Ed.), Intech, 2013 (http://www.intechopen.com/books/polymer-science; 20-12-2013)
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
86 87
is therefore preferred in those applications in which swollen starch granules, or remnants thereof, are unwanted [3].A change in the amylose content of a formulation is expected to affect the influence of amphiphilic additives considerably. The interior of the amylose helix coil is slightly hydrophobic and tends to capture linear hydrocarbon chains [4]. Furthermore, the fine-stranded amylose network tends to aggregate into thicker strands when emulsifiers are added [5]. Starches with a low amylose content are therefore usually recommended when a reduction in ingredients of the formulation (e.g. in latexes applicable as adhesives) is desired. Potato starch has the lowest lipid and protein content and the highest level of bound phosphate groups of the starches available. These properties ensure efficient processing, increase the stability of the formulation obtained and result usually in a final product with the lowest level of reactor fouling during processing. The phosphate groups are, however, mainly present in the amylopectin fraction and solutions based on RP starch are therefore still prone to considerable retrogradation [1]. Unexpected high levels of viscosity can be achieved if WP starch based products are mixed with molecules containing hydrophobic groups (e.g. fatty acids) and this behaviour is not observed (to the same extent) if other waxy varieties are used [6,7]. This is often explained (on a molecular level) by the association between the amylopectin side chains and the hydrophobic groups. The corresponding mixtures display reversible thixotropic (shear thinning and formation of reversible gels) behaviour and this is frequently a desired property in those cases in which a fluid material needs to be applied on a surface (e.g. adhesive, coating, cream, etc) [6-8]. This ability of WP starch to associate with hydrophobic groups makes it a less suitable ingredient for formulations in which small size emulsifiers (i.e. detergents) need to be present in the water phase during preparation or application.Latexes based on octenyl succinic anhydride (OSA) modified starch are already marked as interesting ingredients for making adhesives and paints (Figure 1) [9,10].
Figure 1: The chemical structure of OSA starch.
The combination of the hydrophobic octenyl group and the hydrophilic characteristic of starch confers the OSA starch fragments an amphiphilic nature and renders them “green” alternatives for detergents or emulsifiers. WP starch stabilized latexes are recommended, though a systematic comparison of WP with RP is still lacking in the open literature [11]. This comparative evaluation was the main aim of the present work. Moreover, potato starch with different maximum degrees of substitution (DSmax) of OSA (DSmax 0.01, 0.02, 0.03 and 0.04) was used and the actual need of derivatization was verified by evaluating the unmodified counterparts. The starches were physically modified by extrusion in order to minimize differences in dissolution characteristics between RP and WP starch [12,13]. The impact of different DSmax on the protective colloid characteristics of extruded RP and WP starch was evaluated in a polyvinyl acetate (pVAc) based free radical polymerization (in absence of a detergent, emulsifier and anti-foaming agent) [14,15].
AbstractHydrophobic starches are suitable to replace synthetic protective colloids in free radical polymerizations. Octenyl succinylated waxy potato starch exhibits distinct reversible thixotropic behaviour after dissolution in water and this makes it an interesting protective colloid for polyvinyl acetate based latexes. The effect of degree of substitution (0 to 0.04) on preparation and product characteristics was evaluated for latexes protected with both regular and waxy potato starch. The starch products were extruded before use to introduce a modest degree of degradation and to make them cold-water soluble. The selected polymerization conditions (dry matter = 53 wt %, starch on polyvinyl acetate = 10 wt %, polymerization temperature = 80 °C) allow the use of starch derivatives with a degree of substitution up to 0.02 with respect to coagulation and fouling of the reactor. The substitution degree of potato starch influenced the volume mean diameter (0.6 – 2.9 μm) and the level of polydispersity of the particle size distribution. Latex viscosities in the range of 1 000 – 3 000 mPa·s were achieved with the regular potato starch based products, whilst counterparts based on waxy potato starch range from 1 000 to over 10 000 mPa·s. The increased latex viscosity of the latexes stabilized with octenyl succinate waxy potato starch can be tentatively explained by the associative behaviour between waxy potato starch and the octenyl succinate groups.
IntroductionStarch is found in cereal grains (maize, rice, wheat, barley, oat, sorghum), stems (sago palm), legume seeds (beans, peas), roots (sweet potatoes, cassave, arrow roots, yam) and tubers (potatoes). Considerable structural differences are present between the varieties. D-glucose molecules are bound to each other mostly by α-D-(1,4) linkages and with a α-D-(1,6) bond at a branching point. Differences in polymer chain length, number and structure of branches and triple substituted D-glucose units give rise to a large number of polymer configurations. Starch molecules belong to either the subgroup amylopectin (high degree of polymerization; branched structure) or amylose (lower degree of polymerization; mainly linear chains), with starch granules typically containing 0 to 30% of the latter [1]. The shape and size of the starch granules varies from 1 to 100 μm and their lipid content ranges from 0.1 to 1.2%. Disintegration of starch granules in water requires shear and temperatures in the range of 60 to over 100°C depending on the botanical origin of the starch used. The obtained solution forms a gel upon cooling to room temperature and the degree and speed of this process is strongly correlated with the amylose content and structure. Dissolved amylose chains are present as double helical coils that tend to line up in bundles (retrogradation) and form tightly bound structures, the latter being not easily dissolved in water anymore [1]. The use of amylopectin (commonly referred to as waxy) starch-based products is recommended in applications were the occurrence of retrogradation after dissolution is undesirable (e.g. wallpaper adhesives). The absence of amylose also improves the dissolution characteristics of the starch granule considerably. Amylose containing granules start to swell after the amylose is leached out and it usually takes some time to completely dissolve the granule remnants. On the contrary, waxy potato (WP) starch based granules immediately transform into a macromolecular dispersion after the gelatinization temperature is reached. The gelation temperature of WP starch is slightly higher than that of regular potato (RP) starch but the disintegration of the swollen granules take place at a lower temperature [2]. This type of starch
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
86 87
is therefore preferred in those applications in which swollen starch granules, or remnants thereof, are unwanted [3].A change in the amylose content of a formulation is expected to affect the influence of amphiphilic additives considerably. The interior of the amylose helix coil is slightly hydrophobic and tends to capture linear hydrocarbon chains [4]. Furthermore, the fine-stranded amylose network tends to aggregate into thicker strands when emulsifiers are added [5]. Starches with a low amylose content are therefore usually recommended when a reduction in ingredients of the formulation (e.g. in latexes applicable as adhesives) is desired. Potato starch has the lowest lipid and protein content and the highest level of bound phosphate groups of the starches available. These properties ensure efficient processing, increase the stability of the formulation obtained and result usually in a final product with the lowest level of reactor fouling during processing. The phosphate groups are, however, mainly present in the amylopectin fraction and solutions based on RP starch are therefore still prone to considerable retrogradation [1]. Unexpected high levels of viscosity can be achieved if WP starch based products are mixed with molecules containing hydrophobic groups (e.g. fatty acids) and this behaviour is not observed (to the same extent) if other waxy varieties are used [6,7]. This is often explained (on a molecular level) by the association between the amylopectin side chains and the hydrophobic groups. The corresponding mixtures display reversible thixotropic (shear thinning and formation of reversible gels) behaviour and this is frequently a desired property in those cases in which a fluid material needs to be applied on a surface (e.g. adhesive, coating, cream, etc) [6-8]. This ability of WP starch to associate with hydrophobic groups makes it a less suitable ingredient for formulations in which small size emulsifiers (i.e. detergents) need to be present in the water phase during preparation or application.Latexes based on octenyl succinic anhydride (OSA) modified starch are already marked as interesting ingredients for making adhesives and paints (Figure 1) [9,10].
Figure 1: The chemical structure of OSA starch.
The combination of the hydrophobic octenyl group and the hydrophilic characteristic of starch confers the OSA starch fragments an amphiphilic nature and renders them “green” alternatives for detergents or emulsifiers. WP starch stabilized latexes are recommended, though a systematic comparison of WP with RP is still lacking in the open literature [11]. This comparative evaluation was the main aim of the present work. Moreover, potato starch with different maximum degrees of substitution (DSmax) of OSA (DSmax 0.01, 0.02, 0.03 and 0.04) was used and the actual need of derivatization was verified by evaluating the unmodified counterparts. The starches were physically modified by extrusion in order to minimize differences in dissolution characteristics between RP and WP starch [12,13]. The impact of different DSmax on the protective colloid characteristics of extruded RP and WP starch was evaluated in a polyvinyl acetate (pVAc) based free radical polymerization (in absence of a detergent, emulsifier and anti-foaming agent) [14,15].
AbstractHydrophobic starches are suitable to replace synthetic protective colloids in free radical polymerizations. Octenyl succinylated waxy potato starch exhibits distinct reversible thixotropic behaviour after dissolution in water and this makes it an interesting protective colloid for polyvinyl acetate based latexes. The effect of degree of substitution (0 to 0.04) on preparation and product characteristics was evaluated for latexes protected with both regular and waxy potato starch. The starch products were extruded before use to introduce a modest degree of degradation and to make them cold-water soluble. The selected polymerization conditions (dry matter = 53 wt %, starch on polyvinyl acetate = 10 wt %, polymerization temperature = 80 °C) allow the use of starch derivatives with a degree of substitution up to 0.02 with respect to coagulation and fouling of the reactor. The substitution degree of potato starch influenced the volume mean diameter (0.6 – 2.9 μm) and the level of polydispersity of the particle size distribution. Latex viscosities in the range of 1 000 – 3 000 mPa·s were achieved with the regular potato starch based products, whilst counterparts based on waxy potato starch range from 1 000 to over 10 000 mPa·s. The increased latex viscosity of the latexes stabilized with octenyl succinate waxy potato starch can be tentatively explained by the associative behaviour between waxy potato starch and the octenyl succinate groups.
IntroductionStarch is found in cereal grains (maize, rice, wheat, barley, oat, sorghum), stems (sago palm), legume seeds (beans, peas), roots (sweet potatoes, cassave, arrow roots, yam) and tubers (potatoes). Considerable structural differences are present between the varieties. D-glucose molecules are bound to each other mostly by α-D-(1,4) linkages and with a α-D-(1,6) bond at a branching point. Differences in polymer chain length, number and structure of branches and triple substituted D-glucose units give rise to a large number of polymer configurations. Starch molecules belong to either the subgroup amylopectin (high degree of polymerization; branched structure) or amylose (lower degree of polymerization; mainly linear chains), with starch granules typically containing 0 to 30% of the latter [1]. The shape and size of the starch granules varies from 1 to 100 μm and their lipid content ranges from 0.1 to 1.2%. Disintegration of starch granules in water requires shear and temperatures in the range of 60 to over 100°C depending on the botanical origin of the starch used. The obtained solution forms a gel upon cooling to room temperature and the degree and speed of this process is strongly correlated with the amylose content and structure. Dissolved amylose chains are present as double helical coils that tend to line up in bundles (retrogradation) and form tightly bound structures, the latter being not easily dissolved in water anymore [1]. The use of amylopectin (commonly referred to as waxy) starch-based products is recommended in applications were the occurrence of retrogradation after dissolution is undesirable (e.g. wallpaper adhesives). The absence of amylose also improves the dissolution characteristics of the starch granule considerably. Amylose containing granules start to swell after the amylose is leached out and it usually takes some time to completely dissolve the granule remnants. On the contrary, waxy potato (WP) starch based granules immediately transform into a macromolecular dispersion after the gelatinization temperature is reached. The gelation temperature of WP starch is slightly higher than that of regular potato (RP) starch but the disintegration of the swollen granules take place at a lower temperature [2]. This type of starch
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
88 89
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM dosage was also monitored with a balance. One syringe pump was used to add a premix of SPS and SBC and a second syringe pump was used to add 0.3 M STS after the actual polymerization was finished.
ProcedureA 10 wt % starch mixture was prepared by slowly adding the starch derivative to a demi-water containing beaker while thoroughly mixing (3-bladed impeller; 1000 RPM; 10 minutes). The polymerization reactor was filled with 262.5 g of this mixture and automatic mixing (120 RPM) was started. The applied water bath temperature of the reactor (WTR) and dosage protocols of VAM, SPS/SBC and STS are given in Figure 3. Oxygen removal from the reaction mixture as pre-polymerization step was omitted because a pre-dosage of VAM already results in a considerable reduction in amount of dissolved oxygen. The initial reaction mixture temperature (RMT) is around 76°C at the selected reaction conditions. This is even higher than the boiling point of VAM (72 °C) and the added VAM will therefore immediately shift from the liquid to the gas phase. The phase transition will induce a volume expansion of VAM which will lower the oxygen content of the reactor considerably [15]. VAM was used without inhibitor removal in order to ensure close resemblance to experimental conditions applied at industrial level. A total of 0.27 kg VAM was added in all cases with a pre-dosage level of 15 g. The actual dosage was monitored in time and the amount of VAM added was used for mass balance calculations. 1.5 g SPS and 2.0 g SBC were dissolved together in 46.5 g demineralised water and 36 ml of this mixture was added during the polymerization. The actual addition of the mixture starts after 104 minutes with a pre-dosage of 4.5 ml in 1 minute followed by 31.5 ml with a dosage speed of 5.25 ml/hr. 2.7 ml 0.3 M STS ml was added with 2.7 ml/hr after the WTR dropped significantly below 65°C during cooling down. Agitation was continued for at least one hour after the WTR reaches 20°C. The dispersion was transferred into a storage container without any additional treatments and stored at room temperature.
Figure 3: WTR profile and dosage protocols of SPS/SBC, VAM and STS.
Experimental MaterialsOctenyl succinic anhydride (OSA) from Milliken Chemicals was used to synthesize regular potato (RP) and waxy potato (WP) starch (AVEBE U.A.; Food grade) derivatives with a maximum degree of substitution (DSmax) of 0.01, 0.02, 0.03 and 0.04 on potato starch in suspension at pH 8.5 [16]. This is the optimal pH for this type of derivatization [17]. However, NaOH can also react directly with OSA or saponify the OSA starch formed [18]. The NaOH consumption during processing was therefore used to get an impression of the efficiency of the esterfication. The average efficiency turned out to be 70 % for RP starch and 85 % for WP starch based products. The actual degree of substitution (DSact) is calculated from efficiency and DSmax. The obtained products were extruded (Continua C37; 2 co-rotating axes; 16D with scissor at 10.2; mold: 2 times 3.2 mm; 300 RPM; jacket: 125°C; moisture level 26%; feed: 15 kg/hrs) and subsequently milled (Peppink mill 2.0 mm and 0.5 mm sieve). The vinyl acetate monomer (VAM) was purchased from ACROS and contains 3-30 ppm hydroquinone as inhibitor. Analytical reagent grade sodium persulfate (SPS) was supplied by VWR International. Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) were both of analytical quality and obtained from Merck Germany. STS was added as a 0.3 M solution and SPS and SBC were added together as a mixture in water with 3 % SPS and 4 % SBC on weight in total. All ingredients were used without additional purification. The solvent was demineralised water in all cases.
EquipmentA jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles with a width of 1 cm and an outer dimension of 10.5 x 7 cm (height x diameter)) was applied. A lid made of borosilicate glass with several connection points was placed on top and the reactor was completely insulated with a radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the headspace temperature (HST). The feeding lines of VAM and the SPS/SBC mixture were placed outside the reflux region with the aid of an accessory to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 2).
Figure 2: Schematic representation of the polymerization reactor used.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
88 89
VAM was dosed with a peristaltic pump equipped with polytetrafluoroethylene tubing (4 mm) and the volume removed from the storage bottle was replaced by dry nitrogen. The actual VAM dosage was also monitored with a balance. One syringe pump was used to add a premix of SPS and SBC and a second syringe pump was used to add 0.3 M STS after the actual polymerization was finished.
ProcedureA 10 wt % starch mixture was prepared by slowly adding the starch derivative to a demi-water containing beaker while thoroughly mixing (3-bladed impeller; 1000 RPM; 10 minutes). The polymerization reactor was filled with 262.5 g of this mixture and automatic mixing (120 RPM) was started. The applied water bath temperature of the reactor (WTR) and dosage protocols of VAM, SPS/SBC and STS are given in Figure 3. Oxygen removal from the reaction mixture as pre-polymerization step was omitted because a pre-dosage of VAM already results in a considerable reduction in amount of dissolved oxygen. The initial reaction mixture temperature (RMT) is around 76°C at the selected reaction conditions. This is even higher than the boiling point of VAM (72 °C) and the added VAM will therefore immediately shift from the liquid to the gas phase. The phase transition will induce a volume expansion of VAM which will lower the oxygen content of the reactor considerably [15]. VAM was used without inhibitor removal in order to ensure close resemblance to experimental conditions applied at industrial level. A total of 0.27 kg VAM was added in all cases with a pre-dosage level of 15 g. The actual dosage was monitored in time and the amount of VAM added was used for mass balance calculations. 1.5 g SPS and 2.0 g SBC were dissolved together in 46.5 g demineralised water and 36 ml of this mixture was added during the polymerization. The actual addition of the mixture starts after 104 minutes with a pre-dosage of 4.5 ml in 1 minute followed by 31.5 ml with a dosage speed of 5.25 ml/hr. 2.7 ml 0.3 M STS ml was added with 2.7 ml/hr after the WTR dropped significantly below 65°C during cooling down. Agitation was continued for at least one hour after the WTR reaches 20°C. The dispersion was transferred into a storage container without any additional treatments and stored at room temperature.
Figure 3: WTR profile and dosage protocols of SPS/SBC, VAM and STS.
Experimental MaterialsOctenyl succinic anhydride (OSA) from Milliken Chemicals was used to synthesize regular potato (RP) and waxy potato (WP) starch (AVEBE U.A.; Food grade) derivatives with a maximum degree of substitution (DSmax) of 0.01, 0.02, 0.03 and 0.04 on potato starch in suspension at pH 8.5 [16]. This is the optimal pH for this type of derivatization [17]. However, NaOH can also react directly with OSA or saponify the OSA starch formed [18]. The NaOH consumption during processing was therefore used to get an impression of the efficiency of the esterfication. The average efficiency turned out to be 70 % for RP starch and 85 % for WP starch based products. The actual degree of substitution (DSact) is calculated from efficiency and DSmax. The obtained products were extruded (Continua C37; 2 co-rotating axes; 16D with scissor at 10.2; mold: 2 times 3.2 mm; 300 RPM; jacket: 125°C; moisture level 26%; feed: 15 kg/hrs) and subsequently milled (Peppink mill 2.0 mm and 0.5 mm sieve). The vinyl acetate monomer (VAM) was purchased from ACROS and contains 3-30 ppm hydroquinone as inhibitor. Analytical reagent grade sodium persulfate (SPS) was supplied by VWR International. Sodium bicarbonate (SBC) and sodium thiosulfate pentahydrate (STS) were both of analytical quality and obtained from Merck Germany. STS was added as a 0.3 M solution and SPS and SBC were added together as a mixture in water with 3 % SPS and 4 % SBC on weight in total. All ingredients were used without additional purification. The solvent was demineralised water in all cases.
EquipmentA jacketed stainless steel (316) reactor (1 l) equipped with a stainless steel (316) spiral ribbon stirrer (2 cycles with a width of 1 cm and an outer dimension of 10.5 x 7 cm (height x diameter)) was applied. A lid made of borosilicate glass with several connection points was placed on top and the reactor was completely insulated with a radiator foil. A reflux cooler was placed on top together with a pt-100 probe for measuring the headspace temperature (HST). The feeding lines of VAM and the SPS/SBC mixture were placed outside the reflux region with the aid of an accessory to minimize the contamination of VAM with water and premature dissociation of SPS (Figure 2).
Figure 2: Schematic representation of the polymerization reactor used.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
90 91
Table 1B: Latex (WP) related responses: Viscosity, dry matter and PSD variables VMD and span.
Code OSA derivatization(DS)
Viscosity (mPa·s)
Dry matter(%)
PSD(mm)
Maximum Actual Calculated Found Recovery VMD SpanWO0 0 0 1490 52.8 52.5 99.5 1.6 2.3WO1 0.01 0.008 3950 53.1 51.7 97.4 7.4 2.5WO2 0.02 0.017 11700 53.2 52.5 98.7 4.2 1.9WO3 0.03 0.026 16380 53.0 51.5 97.2 6.9 1.9WO4 0.04 0.034 13520 52.9 51.3 97.0 8.3 1.9
Remark: Calculated dry matter is based on the actual dosed amounts.
The selected formulation (10 wt % starch derivative on pVAc) and polymerization procedure (80 °C; SPS initiation) resulted in a latex with a viscosity of approximately 1500 mPa·s if WP or RP starch was used for stabilization without an OSA derivatization. Application of WP and RP starches with an OSA derivatization resulted in distinct higher viscosities than their unsubstituted counterparts. Moreover, latexes based on WP starch displayed significantly higher viscosities than their counterparts based on RP starch. The level of reactor fouling after the polymerization procedure was also slightly lower for the varieties based on WP starch. This reactor fouling is inversely proportional to the recoveries found in dry matter. This is in line with expectation because the reactor fouling originates mainly from the solid part of the latex. Some small differences in dry matter content were present between WP and RP based products. A part of the observed differences might therefore originate from fluctuations in dry matter content. However, the actual impact of these differences is assumed to be negligible. The latexes stabilized with RO3 and RO4 were excluded from further evaluation due to their distinct lower level of recovery (i.e. < 96 %) since then a relevant proportion of the end product (either starch or the (grafted) polymer pVAc onto starch) would systematically be excluded from the analysis.The use of OSA starch as protective colloid did not only result in higher viscosities but also in an increased VMD of the latex particles. A possible explanation is capture of the VAM by the clusters of hydrophobic moieties of the starch derivative in the initial stage of the polymerization. These VAM rich clusters are excellent nucleation sites for the actual polymerization and the polymer formed is stabilized during this process by the starch derivatives of these clusters as well. The VMD of the WP starch stabilized latexes tended to be lower than their counterparts based on RP starch and this appeared to be also the case for the span of the PSD obtained. The observed differences in viscosity between latexes based on WP and RP starch can therefore be partly deduced to differences in PSD. The DSmax did not correlate with VMD, which in turn did not correlate with the latex viscosity. However, DSmax seemed to correlate with viscosity; thus another factor (influenced by DSmax), besides VMD, must display a relevant influence on the measured viscosity. This is probably the amount of starch derivative dissolved in the water phase. The actual PSD of the varieties based on WP starch (the one for the RP starch based products was very similar and not shown for brevity) are given in Figure 4. The importance of DSmax on PSD as well as the absence of a directly clear correlation is both evident. There was also a pronounced impact of DSmax on the level of bi-modality of the PSD, but the observed behaviour requires additional research in order to obtain a deeper understanding.
CharacterizationViscosity, pH and dry matter were determined with the help of a Brookfield DV-II+( 20 RPM), WTW pH 320 and Mettler Toledo PM100/LP16 (80°C), respectively. Ethanal and residual VAM were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) gas and a flame ionization detector detector. The gas chromatograph measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 min and the supernatant was mixed 1:1 with 5 mM NaOH. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. Particle size distributions (PSD) were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13 - 32.5 μm). Fraunhofer theory based calculations were used and the obtained particle size distributions are ISO 13320 compliant. Glass transition temperatures (Tg) were derived from total heat flow and reversing heat flow curves determined with a modulated differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg dispersion as is).
ResultsScreeningThe procedures described in the experimental section were used to prepare extruded RP and WP starch with different levels of OSA derivatization. The selected DSmax were 0.01, 0.02, 0.03 and 0.04. The efficiency of the OSA reaction is ~70 % for RP and ~85 % for WP starch and the DSact were calculated with these values. The OSA products based on RP starch are designated as RO1, RO2, RO3 and RO4 and the WP starch varieties as WO1, WO2, WO3 and WO4. The starches with the codes RO0 and WO0 represent the extruded RP and WP starches without an OSA derivatization. The ten obtained products were tested as protective colloid in the pVAc polymerization procedure described in the experimental section. The obtained latexes were characterized with the variables viscosity, dry matter, volume mean diameter (VMD) and span (Tabel 1A & 1B).
Table 1A: Latex (RP) related responses: Viscosity, dry matter and PSD variables VMD and span.
Code OSA derivatization(DS)
Viscosity (mPa·s)
Dry matter(%)
PSD(mm)
Maximum Actual Calculated Found Recovery VMD SpanRO0 0 0 1475 52.4 51.6 98.6 2.1 2.4RO1 0.01 0.007 2460 53.0 51.6 97.2 9.8 2.9RO2 0.02 0.014 3100 53.1 51.6 97.2 6.6 2.6RO3 0.03 0.021 9560 52.4 49.3 94.0 4.7 1.7RO4 0.04 0.028 8240 53.0 50.6 95.5 7.4 1.9
Remark: Calculated dry matter is based on the actual dosed amounts.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
90 91
Table 1B: Latex (WP) related responses: Viscosity, dry matter and PSD variables VMD and span.
Code OSA derivatization(DS)
Viscosity (mPa·s)
Dry matter(%)
PSD(mm)
Maximum Actual Calculated Found Recovery VMD SpanWO0 0 0 1490 52.8 52.5 99.5 1.6 2.3WO1 0.01 0.008 3950 53.1 51.7 97.4 7.4 2.5WO2 0.02 0.017 11700 53.2 52.5 98.7 4.2 1.9WO3 0.03 0.026 16380 53.0 51.5 97.2 6.9 1.9WO4 0.04 0.034 13520 52.9 51.3 97.0 8.3 1.9
Remark: Calculated dry matter is based on the actual dosed amounts.
The selected formulation (10 wt % starch derivative on pVAc) and polymerization procedure (80 °C; SPS initiation) resulted in a latex with a viscosity of approximately 1500 mPa·s if WP or RP starch was used for stabilization without an OSA derivatization. Application of WP and RP starches with an OSA derivatization resulted in distinct higher viscosities than their unsubstituted counterparts. Moreover, latexes based on WP starch displayed significantly higher viscosities than their counterparts based on RP starch. The level of reactor fouling after the polymerization procedure was also slightly lower for the varieties based on WP starch. This reactor fouling is inversely proportional to the recoveries found in dry matter. This is in line with expectation because the reactor fouling originates mainly from the solid part of the latex. Some small differences in dry matter content were present between WP and RP based products. A part of the observed differences might therefore originate from fluctuations in dry matter content. However, the actual impact of these differences is assumed to be negligible. The latexes stabilized with RO3 and RO4 were excluded from further evaluation due to their distinct lower level of recovery (i.e. < 96 %) since then a relevant proportion of the end product (either starch or the (grafted) polymer pVAc onto starch) would systematically be excluded from the analysis.The use of OSA starch as protective colloid did not only result in higher viscosities but also in an increased VMD of the latex particles. A possible explanation is capture of the VAM by the clusters of hydrophobic moieties of the starch derivative in the initial stage of the polymerization. These VAM rich clusters are excellent nucleation sites for the actual polymerization and the polymer formed is stabilized during this process by the starch derivatives of these clusters as well. The VMD of the WP starch stabilized latexes tended to be lower than their counterparts based on RP starch and this appeared to be also the case for the span of the PSD obtained. The observed differences in viscosity between latexes based on WP and RP starch can therefore be partly deduced to differences in PSD. The DSmax did not correlate with VMD, which in turn did not correlate with the latex viscosity. However, DSmax seemed to correlate with viscosity; thus another factor (influenced by DSmax), besides VMD, must display a relevant influence on the measured viscosity. This is probably the amount of starch derivative dissolved in the water phase. The actual PSD of the varieties based on WP starch (the one for the RP starch based products was very similar and not shown for brevity) are given in Figure 4. The importance of DSmax on PSD as well as the absence of a directly clear correlation is both evident. There was also a pronounced impact of DSmax on the level of bi-modality of the PSD, but the observed behaviour requires additional research in order to obtain a deeper understanding.
CharacterizationViscosity, pH and dry matter were determined with the help of a Brookfield DV-II+( 20 RPM), WTW pH 320 and Mettler Toledo PM100/LP16 (80°C), respectively. Ethanal and residual VAM were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) gas and a flame ionization detector detector. The gas chromatograph measurement was performed on water diluted dispersions (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 min and the supernatant was mixed 1:1 with 5 mM NaOH. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. Particle size distributions (PSD) were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (Range: 0.13 - 32.5 μm). Fraunhofer theory based calculations were used and the obtained particle size distributions are ISO 13320 compliant. Glass transition temperatures (Tg) were derived from total heat flow and reversing heat flow curves determined with a modulated differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1 °C/min; amplitude: 0.5 °C; period: 60 s; large volume stainless steel pans; 20-50 mg dispersion as is).
ResultsScreeningThe procedures described in the experimental section were used to prepare extruded RP and WP starch with different levels of OSA derivatization. The selected DSmax were 0.01, 0.02, 0.03 and 0.04. The efficiency of the OSA reaction is ~70 % for RP and ~85 % for WP starch and the DSact were calculated with these values. The OSA products based on RP starch are designated as RO1, RO2, RO3 and RO4 and the WP starch varieties as WO1, WO2, WO3 and WO4. The starches with the codes RO0 and WO0 represent the extruded RP and WP starches without an OSA derivatization. The ten obtained products were tested as protective colloid in the pVAc polymerization procedure described in the experimental section. The obtained latexes were characterized with the variables viscosity, dry matter, volume mean diameter (VMD) and span (Tabel 1A & 1B).
Table 1A: Latex (RP) related responses: Viscosity, dry matter and PSD variables VMD and span.
Code OSA derivatization(DS)
Viscosity (mPa·s)
Dry matter(%)
PSD(mm)
Maximum Actual Calculated Found Recovery VMD SpanRO0 0 0 1475 52.4 51.6 98.6 2.1 2.4RO1 0.01 0.007 2460 53.0 51.6 97.2 9.8 2.9RO2 0.02 0.014 3100 53.1 51.6 97.2 6.6 2.6RO3 0.03 0.021 9560 52.4 49.3 94.0 4.7 1.7RO4 0.04 0.028 8240 53.0 50.6 95.5 7.4 1.9
Remark: Calculated dry matter is based on the actual dosed amounts.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
92 93
The observed behaviour is not micelle related because the WO0 formulation did not contain any ingredients with an amphiphilic nature. However, an interaction between pVAc generated during polymerization and WP starch might by feasible since WP starch is known to form viscosity increasing aggregates with hydrophobic molecules [6,7]. Distinct changes in torque profile were observed between 1.5 and 4.5 hours if WO1, WO2, WO3 or WO4 was used for stabilization instead of WO0. Particle formation takes place in this period of time according a visual observation of the same polymerizations executed in a transparent reactor made of borosilicate glass. The level of torque is expected to depend on the level of excess VAM because the size of a latex particle is proportional to its VAM content [19,20]. The observed increases in torque between 1.5 and 4.5 hrs of processing are therefore not only based on the PSD but on free VAM content as well. The decay in torque after the VAM dosage was stopped confirms the significance of this assumption. The monomer stored in particles is consumed from this point on and results in a considerable torque reduction in time. The stirrer torque profiles of the latexes stabilized with starch derivatives WO3 and WO4 display a higher level of erratic behaviour than the other three varieties. The fluctuations in stirrer torque probably originated from depositions of polymer on the stirrer blade during processing. This is plausible because the amount of depositions on the stirrer blade was proportional with the level of derivatization.Figure 4 shows that the torque increase is proportional to the DSmax of the starch once the monomer dosage is finished. The maximum torque during the preparation of the latex stabilized with WO1 was higher than its WO0 counterpart. The observed behaviour is probably induced by a change in the particle formation process because the corresponding latex does not only contain larger particles but has an increased span as well (Table 1). This assumption is also in line with the observed changes in torque profile if WO2 is used for stabilization. A change from WO1 to WO2 results in a reduction in span and VMD of the PSD. Moreover, an increase in initial torque was observed if WO3 was used instead of WO2 and this replacement induced a change in PSD as well. The VMD showed a pronounced increase after this change and this is indicative that the PSD is not the only factor that controls the torque profile during processing. The fact that WO2 and WO3 have approximately the same span is also in agreement with this hypothesis. The PSD appeared to be linked to the level of torque in the early stage of processing because latexes based on WO3. WO4 had a similar PSD and the corresponding torque profiles were the same up to 3 hrs of processing. The observed differences in latex particles smaller than 3 μm and larger 20 μm limits the validity of this hypothesis, even if these deviations might also originated after 3 hrs of processing
Potential wood adhesivesLatexes stabilized with WO2 and RO2 were selected for a more elaborate investigation because of their lower level of reactor fouling and the fact that the corresponding latex viscosities range from 3 000 to over 10 000 mPa·s (Table 1). This range coincides largely with the viscosity specification of roller applications, which is frequently used in the wood adhesive industry [21]. Latexes stabilized with RO2 and WO2 were prepared and compared to the WO0, whose behaviour was similar in all aspects to RO0 (which was then, for the sake of brevity, left out of the discussion). The polymerizations were executed in triplicate and the
Figure 4: PSD’s of the latexes WO0, WO1, WO2, WO3 and WO4.
During the emulsion polymerization process, the latex based on WO0 showed a gradual increase in torque of the stirrer once the VAM dosage was started. This tended to level off at the point when the initiator solution was dosed and the VAM dosage was reduced to a third (Figure 5).
Figure 5: Stirrer torque as function of the polymerization time for the latexes WO0, WO1, WO2, WO3 and WO4. A moving average is used for smoothing and the lines are aligned by reducing each measuring point by the 1.5 hrs value of the corresponding data set.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
92 93
The observed behaviour is not micelle related because the WO0 formulation did not contain any ingredients with an amphiphilic nature. However, an interaction between pVAc generated during polymerization and WP starch might by feasible since WP starch is known to form viscosity increasing aggregates with hydrophobic molecules [6,7]. Distinct changes in torque profile were observed between 1.5 and 4.5 hours if WO1, WO2, WO3 or WO4 was used for stabilization instead of WO0. Particle formation takes place in this period of time according a visual observation of the same polymerizations executed in a transparent reactor made of borosilicate glass. The level of torque is expected to depend on the level of excess VAM because the size of a latex particle is proportional to its VAM content [19,20]. The observed increases in torque between 1.5 and 4.5 hrs of processing are therefore not only based on the PSD but on free VAM content as well. The decay in torque after the VAM dosage was stopped confirms the significance of this assumption. The monomer stored in particles is consumed from this point on and results in a considerable torque reduction in time. The stirrer torque profiles of the latexes stabilized with starch derivatives WO3 and WO4 display a higher level of erratic behaviour than the other three varieties. The fluctuations in stirrer torque probably originated from depositions of polymer on the stirrer blade during processing. This is plausible because the amount of depositions on the stirrer blade was proportional with the level of derivatization.Figure 4 shows that the torque increase is proportional to the DSmax of the starch once the monomer dosage is finished. The maximum torque during the preparation of the latex stabilized with WO1 was higher than its WO0 counterpart. The observed behaviour is probably induced by a change in the particle formation process because the corresponding latex does not only contain larger particles but has an increased span as well (Table 1). This assumption is also in line with the observed changes in torque profile if WO2 is used for stabilization. A change from WO1 to WO2 results in a reduction in span and VMD of the PSD. Moreover, an increase in initial torque was observed if WO3 was used instead of WO2 and this replacement induced a change in PSD as well. The VMD showed a pronounced increase after this change and this is indicative that the PSD is not the only factor that controls the torque profile during processing. The fact that WO2 and WO3 have approximately the same span is also in agreement with this hypothesis. The PSD appeared to be linked to the level of torque in the early stage of processing because latexes based on WO3. WO4 had a similar PSD and the corresponding torque profiles were the same up to 3 hrs of processing. The observed differences in latex particles smaller than 3 μm and larger 20 μm limits the validity of this hypothesis, even if these deviations might also originated after 3 hrs of processing
Potential wood adhesivesLatexes stabilized with WO2 and RO2 were selected for a more elaborate investigation because of their lower level of reactor fouling and the fact that the corresponding latex viscosities range from 3 000 to over 10 000 mPa·s (Table 1). This range coincides largely with the viscosity specification of roller applications, which is frequently used in the wood adhesive industry [21]. Latexes stabilized with RO2 and WO2 were prepared and compared to the WO0, whose behaviour was similar in all aspects to RO0 (which was then, for the sake of brevity, left out of the discussion). The polymerizations were executed in triplicate and the
Figure 4: PSD’s of the latexes WO0, WO1, WO2, WO3 and WO4.
During the emulsion polymerization process, the latex based on WO0 showed a gradual increase in torque of the stirrer once the VAM dosage was started. This tended to level off at the point when the initiator solution was dosed and the VAM dosage was reduced to a third (Figure 5).
Figure 5: Stirrer torque as function of the polymerization time for the latexes WO0, WO1, WO2, WO3 and WO4. A moving average is used for smoothing and the lines are aligned by reducing each measuring point by the 1.5 hrs value of the corresponding data set.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
94 95
Figure 7: Torque as function of the polymerization time for WO2 and RO2. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate. The dotted line represents the applied WTR profile.
Figure 8: PSD’s as function of the polymerization time with WO2, RO2 and WO0. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate measurement.
Table 2. Product characteristics: VMD, thermal transitions and viscosity.Code Viscosity Thermal transitions (Reversing heat flow) PSD
(mPa·s)
Tg,onset (°C)
Tg,inflection (°C)
Tg,endset (°C)
DTg (°C)
VMD(mm)
WO2 10380 8.8 11.6 14.6 5.8 4.5RO2 3020 8.8 11.0 13.3 4.5 5.3WO0 1280 10.7 13.3 15.3 4.6 1.4
We start by noticing how the thermal properties (Tg,x values in Table 2) of RO2 and WO2 are factually similar to each other with differences within the experimental detection limits (i.e. 0.5 °C). On the other hand relevant variations can be observed in the variables visosity and VMD.
obtained results were averaged. The torque profiles of WO2 and WO0 were significantly different according the variation (2 times the standard deviation (σ)) of the triplicates executed (Figure 6). Both profiles showed a decrease in level of torque after the VAM dosage was finished. This was related to the amount of VAM in the swollen latex particles (vide supra).
Figure 6: Torque as function of the polymerization time for WO2 and WO0. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate. The dotted line represents the dosage profile of VAM.
The difference between WO2 and RO2 was pronounced and close to systematic but only up to 7-8 hours after the initiation of the procedure (Figure 7). From this point the torque of WO2 remained close to constant whilst the torque of RO2 continued to decrease. The difference between WO2 and RO2 became even more pronounced once the WTR dropped below 65°C. Dissolved WP starch is known to associate with hydrophobic groups and this interaction can considerably increase the viscosity of the corresponding mixtures [6,7]. WO2 had a pronounced higher torque level during processing than its RO2 counterpart and it is tempting to assign the increased torque level of WO2 totally to the interaction between WP starch and hydrophobic groups. However, there was also a distinct difference between the DSact of WO2 (0.017) and RO2 (0.014) (Table 1). The actual impact of this difference in DSact is not known and the observed behaviour is probably a combination of a difference in DSact and the association between WP starch and the hydrophobic groups bound to WP starch. The use of OSA potato starch did not only have a distinct impact on the torque profile during the polymerization reaction, but also viscosity level, thermal transitions and PSD of the final products were influenced (Figure 8 and Table 2).
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
94 95
Figure 7: Torque as function of the polymerization time for WO2 and RO2. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate. The dotted line represents the applied WTR profile.
Figure 8: PSD’s as function of the polymerization time with WO2, RO2 and WO0. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate measurement.
Table 2. Product characteristics: VMD, thermal transitions and viscosity.Code Viscosity Thermal transitions (Reversing heat flow) PSD
(mPa·s)
Tg,onset (°C)
Tg,inflection (°C)
Tg,endset (°C)
DTg (°C)
VMD(mm)
WO2 10380 8.8 11.6 14.6 5.8 4.5RO2 3020 8.8 11.0 13.3 4.5 5.3WO0 1280 10.7 13.3 15.3 4.6 1.4
We start by noticing how the thermal properties (Tg,x values in Table 2) of RO2 and WO2 are factually similar to each other with differences within the experimental detection limits (i.e. 0.5 °C). On the other hand relevant variations can be observed in the variables visosity and VMD.
obtained results were averaged. The torque profiles of WO2 and WO0 were significantly different according the variation (2 times the standard deviation (σ)) of the triplicates executed (Figure 6). Both profiles showed a decrease in level of torque after the VAM dosage was finished. This was related to the amount of VAM in the swollen latex particles (vide supra).
Figure 6: Torque as function of the polymerization time for WO2 and WO0. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate. The dotted line represents the dosage profile of VAM.
The difference between WO2 and RO2 was pronounced and close to systematic but only up to 7-8 hours after the initiation of the procedure (Figure 7). From this point the torque of WO2 remained close to constant whilst the torque of RO2 continued to decrease. The difference between WO2 and RO2 became even more pronounced once the WTR dropped below 65°C. Dissolved WP starch is known to associate with hydrophobic groups and this interaction can considerably increase the viscosity of the corresponding mixtures [6,7]. WO2 had a pronounced higher torque level during processing than its RO2 counterpart and it is tempting to assign the increased torque level of WO2 totally to the interaction between WP starch and hydrophobic groups. However, there was also a distinct difference between the DSact of WO2 (0.017) and RO2 (0.014) (Table 1). The actual impact of this difference in DSact is not known and the observed behaviour is probably a combination of a difference in DSact and the association between WP starch and the hydrophobic groups bound to WP starch. The use of OSA potato starch did not only have a distinct impact on the torque profile during the polymerization reaction, but also viscosity level, thermal transitions and PSD of the final products were influenced (Figure 8 and Table 2).
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
96 97
The difference in Tg,inflection between WO0 (13.3 °C) and WO2 (11.6 °) can be explained by either a change in composition or level of branching of the pVAc generated. Octenyl succinylated starch anchored on the surface of the latex particle affects the purity level of the pVAc prepared but the observed difference can probably mainly be explained by a considerable difference in HST during processing (Figure 10).
Figure 10: HST as function of the polymerization time for WO2 and WO0. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate. The dotted line shows the dosage profile of monomer.
The HST during polymerization of WO0 was in the range of 65-68°C. This is close to the boiling point of the VAM-water azeotrope (66°C) and transport of considerable amounts of heat from the reaction mixture to the cooler are therefore feasible [15,26]. The higher water bath power consumption (WPC) of the processing procedure of WO0 with respect to the other two varieties is also in line with this assumption (Figure 11).
Figure 11: WPC as function of the polymerization time with WO2, RO2 and WO0. The lines are smoothed by a moving average and based on a triplicate experiment.
Differences in PSD can influence the level of viscosity and the differences between the PSD of WO0 and WO2 could therefore be a plausible explanation of the observed levels of viscosity [22]. However, the PSD’s of WO2 and RO2 were very similar whilst their viscosities differ considerably. The observed differences in viscosity between WO2 and RO2 are probably related to differences in rheology of the water phase (vide supra). The origin of this behaviour might be attributed to the fact that only WP starch can exhibit associative behaviour with hydrophobic domains [6,7]. Unfortunately, the viscosities at 20°C of the 10 wt % solutions of RO2 (920mPa·s; turbid), WO2 (338 mPa·s; clear) and WO0 (178 mPa·s; clear) cannot be used to confirm this assumption because RO2 showed severe signs of retrogradated material which also resulted in an increase in viscosity. The PSD’s of WO2 and RO2 showed a bimodal distribution and this might be explained by either the occurrence of coagulation of latex particles and/or of two separate nucleation stages during preparation. Both processes can be influenced by differences in the amount of octenyl succinate groups bound to the starch fragments. In the first case, insufficient starch is present at the surface of the the latex particle to prevent it from coalesce. In the other, not enough micelle like structures are present to store the amount of monomer to prevent other nucleation mechanisms (e.g homogeneous nucleation). The latter might also be the reason of the bimodal distribution of the latex stabilized with WO1 (Figure 4).WO2 and RO2 had HST during processing which were sometimes considerable higher than equilibrium HST, especially the WO2-based polymerization (Figure 9). Previous research showed that RMT correlates with the HST [15]. A HST higher than the equilibrium HST is therefore indicative for a RMT higher than generated by the applied WTR. This phenomenon is possible due to the highly exothermic nature of the vinyl based polymerization [23]. The level of pVAc branching introduced during polymerization is proportional to the RMT [24,25]. The RMT of WO2 was systematically higher than RO2 during polymerization but the impact in the level of pVAc branching between both varieties was limited according the Tg’s measured (Table 2).
Figure 9: HST as function of the polymerization time for WO2 and RO2. The lines are smoothed by a moving average and based on a triplicate measurement. The dotted line shows the temperature profile and both straight lines represent the initial equilibrium HST of WO2 and RO2.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
96 97
The difference in Tg,inflection between WO0 (13.3 °C) and WO2 (11.6 °) can be explained by either a change in composition or level of branching of the pVAc generated. Octenyl succinylated starch anchored on the surface of the latex particle affects the purity level of the pVAc prepared but the observed difference can probably mainly be explained by a considerable difference in HST during processing (Figure 10).
Figure 10: HST as function of the polymerization time for WO2 and WO0. The lines are smoothed by a moving average and the error bars represent 2 times σ of a triplicate. The dotted line shows the dosage profile of monomer.
The HST during polymerization of WO0 was in the range of 65-68°C. This is close to the boiling point of the VAM-water azeotrope (66°C) and transport of considerable amounts of heat from the reaction mixture to the cooler are therefore feasible [15,26]. The higher water bath power consumption (WPC) of the processing procedure of WO0 with respect to the other two varieties is also in line with this assumption (Figure 11).
Figure 11: WPC as function of the polymerization time with WO2, RO2 and WO0. The lines are smoothed by a moving average and based on a triplicate experiment.
Differences in PSD can influence the level of viscosity and the differences between the PSD of WO0 and WO2 could therefore be a plausible explanation of the observed levels of viscosity [22]. However, the PSD’s of WO2 and RO2 were very similar whilst their viscosities differ considerably. The observed differences in viscosity between WO2 and RO2 are probably related to differences in rheology of the water phase (vide supra). The origin of this behaviour might be attributed to the fact that only WP starch can exhibit associative behaviour with hydrophobic domains [6,7]. Unfortunately, the viscosities at 20°C of the 10 wt % solutions of RO2 (920mPa·s; turbid), WO2 (338 mPa·s; clear) and WO0 (178 mPa·s; clear) cannot be used to confirm this assumption because RO2 showed severe signs of retrogradated material which also resulted in an increase in viscosity. The PSD’s of WO2 and RO2 showed a bimodal distribution and this might be explained by either the occurrence of coagulation of latex particles and/or of two separate nucleation stages during preparation. Both processes can be influenced by differences in the amount of octenyl succinate groups bound to the starch fragments. In the first case, insufficient starch is present at the surface of the the latex particle to prevent it from coalesce. In the other, not enough micelle like structures are present to store the amount of monomer to prevent other nucleation mechanisms (e.g homogeneous nucleation). The latter might also be the reason of the bimodal distribution of the latex stabilized with WO1 (Figure 4).WO2 and RO2 had HST during processing which were sometimes considerable higher than equilibrium HST, especially the WO2-based polymerization (Figure 9). Previous research showed that RMT correlates with the HST [15]. A HST higher than the equilibrium HST is therefore indicative for a RMT higher than generated by the applied WTR. This phenomenon is possible due to the highly exothermic nature of the vinyl based polymerization [23]. The level of pVAc branching introduced during polymerization is proportional to the RMT [24,25]. The RMT of WO2 was systematically higher than RO2 during polymerization but the impact in the level of pVAc branching between both varieties was limited according the Tg’s measured (Table 2).
Figure 9: HST as function of the polymerization time for WO2 and RO2. The lines are smoothed by a moving average and based on a triplicate measurement. The dotted line shows the temperature profile and both straight lines represent the initial equilibrium HST of WO2 and RO2.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
98 99
STS : Sodium thiosulfate.HST : Headspace temperature of the reactor.WTR : Water bath temperature reactor.RMT : Reaction mixture temperature.PSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulated differential scanning calorimeter.WO0-4 : Extruded WP starch with a DSmax OSA of 0, 0.01, 0.02, 0.03 and 0.04.RO0-4 : Extruded RP starch with a DSmax OSA of 0, 0.01, 0.02, 0.03 and 0.04.VMD : Volume Mean Diameter.σ : Standard deviation.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.WPC : Water bath power consumption.
The HST was higher than the reflux temperature of the azetrope VAM - water (66 °C) most of the time. The optimal azeotrope composition (VAM : water = 93 : 7) was probably never present during the polymerization of WO2, RO2 and WO0, maybe besides the two short periods of 66 °C during the preparation of WO0. The actual RMT was therefore probably much higher than 66 °C most of the time but there were no indications that RMT exceeded the WTR setting of 80 °C. The level of pVAc branching is proportional to RMT and this might be an explanation for the observed inverse correlation between HST and Tg,inflection (vide supra). Radicals were generated by thermal dissociation of SPS and changes in reaction temperature will have an impact on the radical formation process as well. Differences in concentrations of hydrogen ion, ethanal, VAM, acetate, sulfate and thiosulfate in the final product were present but on a limited scale only. Significant correlations between these radical formation related responses and the observed change in latex viscosity and PSD were therefore not expected.
ConclusionsOctenyl succinylation of starch has a considerable impact on the protective colloid properties of the product in a VAM based free radical polymerization. A product with a DSmax 0.02 showed pronounced differences in polymerization process (e.g. torque) and product (viscosity, particle size etc…) characteristics compared to product without this dervatization. The derivatization reduced the energy demand during processing and the latexes showed a higher viscosity and larger particle sizes than counterparts without this type of modification. The latex viscosity appeared to be proportional to the DSmax of the derivatization up to a level of 0.02 for both regular and waxy potato starch based products. The latexes based regular potato starch with a DSmax 0.03 and 0.04 cannot be properly compared with the other products because of a higher loss of material due to coagulation and fouling of the reactor. A change in process conditions is needed to increase the material recovery of this type of latexes and make them more suitable for comparison with latexes based on starch derivatives with a lower degree of substitution. The higher latex viscosity of the waxy potato variety with respect to its regular counterpart can probably be explained by associative behaviour between waxy potato starch fragments and the hydrophobic domains present. This is a unique property of waxy potato starch and fragments of this type of starch apparently still exhibit this characteristic after the applied polymerization conditions.
AbbreviationsWP : Waxy potato starch.RP : Regular potato starch.OSA : Octenyl succinic anhydride.DSmax : Maximum degree of substitution based on the amount of reagent added. pVAc : Polyvinyl acetate.DSact : Actual degree of substitution calculated from DSmax and efficieny.RPM : Revolutions per minute.VAM : Vinyl acetate monomer.SPS : Sodium persulfate.SBC : Sodium bicarbonate.
Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes
98 99
STS : Sodium thiosulfate.HST : Headspace temperature of the reactor.WTR : Water bath temperature reactor.RMT : Reaction mixture temperature.PSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulated differential scanning calorimeter.WO0-4 : Extruded WP starch with a DSmax OSA of 0, 0.01, 0.02, 0.03 and 0.04.RO0-4 : Extruded RP starch with a DSmax OSA of 0, 0.01, 0.02, 0.03 and 0.04.VMD : Volume Mean Diameter.σ : Standard deviation.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.WPC : Water bath power consumption.
The HST was higher than the reflux temperature of the azetrope VAM - water (66 °C) most of the time. The optimal azeotrope composition (VAM : water = 93 : 7) was probably never present during the polymerization of WO2, RO2 and WO0, maybe besides the two short periods of 66 °C during the preparation of WO0. The actual RMT was therefore probably much higher than 66 °C most of the time but there were no indications that RMT exceeded the WTR setting of 80 °C. The level of pVAc branching is proportional to RMT and this might be an explanation for the observed inverse correlation between HST and Tg,inflection (vide supra). Radicals were generated by thermal dissociation of SPS and changes in reaction temperature will have an impact on the radical formation process as well. Differences in concentrations of hydrogen ion, ethanal, VAM, acetate, sulfate and thiosulfate in the final product were present but on a limited scale only. Significant correlations between these radical formation related responses and the observed change in latex viscosity and PSD were therefore not expected.
ConclusionsOctenyl succinylation of starch has a considerable impact on the protective colloid properties of the product in a VAM based free radical polymerization. A product with a DSmax 0.02 showed pronounced differences in polymerization process (e.g. torque) and product (viscosity, particle size etc…) characteristics compared to product without this dervatization. The derivatization reduced the energy demand during processing and the latexes showed a higher viscosity and larger particle sizes than counterparts without this type of modification. The latex viscosity appeared to be proportional to the DSmax of the derivatization up to a level of 0.02 for both regular and waxy potato starch based products. The latexes based regular potato starch with a DSmax 0.03 and 0.04 cannot be properly compared with the other products because of a higher loss of material due to coagulation and fouling of the reactor. A change in process conditions is needed to increase the material recovery of this type of latexes and make them more suitable for comparison with latexes based on starch derivatives with a lower degree of substitution. The higher latex viscosity of the waxy potato variety with respect to its regular counterpart can probably be explained by associative behaviour between waxy potato starch fragments and the hydrophobic domains present. This is a unique property of waxy potato starch and fragments of this type of starch apparently still exhibit this characteristic after the applied polymerization conditions.
AbbreviationsWP : Waxy potato starch.RP : Regular potato starch.OSA : Octenyl succinic anhydride.DSmax : Maximum degree of substitution based on the amount of reagent added. pVAc : Polyvinyl acetate.DSact : Actual degree of substitution calculated from DSmax and efficieny.RPM : Revolutions per minute.VAM : Vinyl acetate monomer.SPS : Sodium persulfate.SBC : Sodium bicarbonate.
100
CHAPTER 6The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
References[1] J.J.M. Swinkels, Composition and properties of commercial native starches, Starch, 37 (1) (1987) 1-5.[2] K. Svegmark, K. Helmersson, G. Nilsson, P.O. Nilsson, R Andersson, E. Svensson, Comparison of potato
amylopectin starches and potato starches – influence of year and varieties, Carbohydrate Polymers, 47 (2002) 331-340.
[3] A.M. Hermansson, K. Svegmark, Developments in the understanding of starch functionality, Trends in food science & technology, 7 (1996) 345
[4] K.A. Murdoch, The amylose-iodine complex, Carbohydrate Research, 233 (1992) 161-174[5] G. Richardson, S. Kidman, M. Langton,A.M. Hermansson, Differences in amylose aggregation and starch gel
formation with emulsifiers, Carbohydrate Polymers 58 (2004) 7–13[6] R.H. Huizenga, H.G. Alrich, I.L. De Groot, Aqueous compositions comprising amylopectin-potato starch and
process for their manufacture (1997) EP799837 [7] P.L. Buwalda, R.P.W. Kesselmans, A.A.M. Maas, H.H. Simonides, Hydrophobic starch derivatives (2000)
EP1141030[8] H.A. Barnes, Thixotropy: A review, J. Non-Newtonian Fluid Mech., 70 (1997) 1-33[9] C.P. Iovine, Y.J. Shih, P.R. Mudge, P.T. Trzasko, Hydrophobically modified starch stabilized vinyl ester polymer
emulsions (1986) EP0223145[10] R.L. Billmers, R. Farwaha, G.S. Yearwood, L. Phan, Thixotropic paint compositions containing hydrophobic
starch derivatives (1999) US6001927 [11] M. Pfalz, M.W. Marnik, D. Gruel (1999) Emulsion polymerization method, EP1232189. [12] J. N. BeMiller, R.L. Whistler: Starch: Chemistry and Technology, third edition, Food science and technology,
International series, Academic Press, Burlington, 2009[13] R.M. Van den Einde, A.J. Van der Goot, R.M. Boom, Understanding molecular weight reduction of starch during
heating-shearing processes, Journal of food science, 68 (8) (2003) 2396-2404[14] K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres,
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects, To be published.
[15] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics, To be published.
[16] H.J. De Vries, C. Semeijn, P.L. Buwalda, Emulsifier (2005) EP1743693[17] Y. Jeon, A. Viswanathan, R.A. Gross, Studies of Starch Esterification: Reactions with Alkenyl succinates in
Aqueous Slurry Systems. Starch/Starke, 51(2-3) (1999) 90-93.[18] G.Q. He, X.Y. Song, H. Ruan, F Chen, Octenyl succinic anhydride modified early indica rice starches differing
in amylose content. Journal of Agricultural and Food Chemistry, 54 (7) (2006) 2775-2779.[19] W.J. Priest, Particle growth in the aqueous polymerization of vinyl acetate, J. Phys. Chem, 56 (1952) 1077-
1082.[20] A. Klein, C.H. Kuist, V.T. Stannet, Vinyl acetate emulsion polymerization. I. Effect of ionic strength and
temperature on monomer solubility in the ionically stabilized polymer particle, J. Polym. Sci. Chem. Ed. 11 (1973) 2111-2126.
[21] P. Cognard, Adhesives and sealants: General knowledge, Application Technique, New curing techniques, Elsevier Ltd, Oxford, 2006
[22] A. Guyot, F. Chu, M. Schneider, C. Graillat, T.F. McKenna, High solid content latexes, Prog Polym Sci, 27 (2002) 1573-1615
[23] H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).
[24] D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837.
[25] P. Cognard, Adhesive bonding of wood and wood based products Part 3: Different types of adhesives, glues and binders used in woodworking, 2005, (http://www.omnexus4adhesives.com; 27-03-2013)
[26] R.L. Zollars, C.T. Chen, D.A. Jones, Distribution of volatile species in a refluxing polymer colloid, AIChE. J., 34 (5) (1988)
100
CHAPTER 6The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
References[1] J.J.M. Swinkels, Composition and properties of commercial native starches, Starch, 37 (1) (1987) 1-5.[2] K. Svegmark, K. Helmersson, G. Nilsson, P.O. Nilsson, R Andersson, E. Svensson, Comparison of potato
amylopectin starches and potato starches – influence of year and varieties, Carbohydrate Polymers, 47 (2002) 331-340.
[3] A.M. Hermansson, K. Svegmark, Developments in the understanding of starch functionality, Trends in food science & technology, 7 (1996) 345
[4] K.A. Murdoch, The amylose-iodine complex, Carbohydrate Research, 233 (1992) 161-174[5] G. Richardson, S. Kidman, M. Langton,A.M. Hermansson, Differences in amylose aggregation and starch gel
formation with emulsifiers, Carbohydrate Polymers 58 (2004) 7–13[6] R.H. Huizenga, H.G. Alrich, I.L. De Groot, Aqueous compositions comprising amylopectin-potato starch and
process for their manufacture (1997) EP799837 [7] P.L. Buwalda, R.P.W. Kesselmans, A.A.M. Maas, H.H. Simonides, Hydrophobic starch derivatives (2000)
EP1141030[8] H.A. Barnes, Thixotropy: A review, J. Non-Newtonian Fluid Mech., 70 (1997) 1-33[9] C.P. Iovine, Y.J. Shih, P.R. Mudge, P.T. Trzasko, Hydrophobically modified starch stabilized vinyl ester polymer
emulsions (1986) EP0223145[10] R.L. Billmers, R. Farwaha, G.S. Yearwood, L. Phan, Thixotropic paint compositions containing hydrophobic
starch derivatives (1999) US6001927 [11] M. Pfalz, M.W. Marnik, D. Gruel (1999) Emulsion polymerization method, EP1232189. [12] J. N. BeMiller, R.L. Whistler: Starch: Chemistry and Technology, third edition, Food science and technology,
International series, Academic Press, Burlington, 2009[13] R.M. Van den Einde, A.J. Van der Goot, R.M. Boom, Understanding molecular weight reduction of starch during
heating-shearing processes, Journal of food science, 68 (8) (2003) 2396-2404[14] K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres,
Modified waxy potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects, To be published.
[15] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics, To be published.
[16] H.J. De Vries, C. Semeijn, P.L. Buwalda, Emulsifier (2005) EP1743693[17] Y. Jeon, A. Viswanathan, R.A. Gross, Studies of Starch Esterification: Reactions with Alkenyl succinates in
Aqueous Slurry Systems. Starch/Starke, 51(2-3) (1999) 90-93.[18] G.Q. He, X.Y. Song, H. Ruan, F Chen, Octenyl succinic anhydride modified early indica rice starches differing
in amylose content. Journal of Agricultural and Food Chemistry, 54 (7) (2006) 2775-2779.[19] W.J. Priest, Particle growth in the aqueous polymerization of vinyl acetate, J. Phys. Chem, 56 (1952) 1077-
1082.[20] A. Klein, C.H. Kuist, V.T. Stannet, Vinyl acetate emulsion polymerization. I. Effect of ionic strength and
temperature on monomer solubility in the ionically stabilized polymer particle, J. Polym. Sci. Chem. Ed. 11 (1973) 2111-2126.
[21] P. Cognard, Adhesives and sealants: General knowledge, Application Technique, New curing techniques, Elsevier Ltd, Oxford, 2006
[22] A. Guyot, F. Chu, M. Schneider, C. Graillat, T.F. McKenna, High solid content latexes, Prog Polym Sci, 27 (2002) 1573-1615
[23] H. De Bruyn, The emulsion polymerization of vinyl acetate, (http://ses.library.usyd.edu.au/bitstream/2123/381/3/adt-NU1999.0006whole.pdf; 20-06-2012).
[24] D. Britton, F. Heatley, P.A. Lovell, Chain transfer to polymer in free-radical bulk and emulsion polymerization of viny acetate studied by NMR spectroscopy, Macromolecules, 31 (1998) 2828-2837.
[25] P. Cognard, Adhesive bonding of wood and wood based products Part 3: Different types of adhesives, glues and binders used in woodworking, 2005, (http://www.omnexus4adhesives.com; 27-03-2013)
[26] R.L. Zollars, C.T. Chen, D.A. Jones, Distribution of volatile species in a refluxing polymer colloid, AIChE. J., 34 (5) (1988)
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
102 103
etc…) made of wood or derived timber products. Polyvinyl acetate (pVAc) based adhesives are usually classified as D1 (for use in dry and temperate climates including indoor furniture) or D2 (for use in relative damp indoor, e.g. kitchen furniture). However, some special formulations can even meet D3 (application in damp indoor or outdoor environment) and D4 (water resistant) [5-6]. This set of characteristics and wide applicability spectrum renders these adhesives particularly attractive for further developments.Spray guns, rollers and trowels are frequently used in woodworking and each application technique requires a different level of viscosity. Viscosities in the range of 300 -1 000 mPa·s are suitable for spray application whilst rollers require values in the range of several thousands mPa·s [1,2]. A target value of at least 20 000 mPa·s is required if a trowel is used for application [1,2]. Shear thinning behaviour with a reversible nature (pseudoplasticity or thixotropy) is desired because this improves the workability significantly [7]. This type of adhesive cannot only be added to a substrate more easily but the chance of dripping is reduced as well [7]. Other important application related parameters are pot life, open time, wet tack, curing time and bond strength [1,2]. Additives (plasticizers, fillers, solvents, thickeners and tackifiers) are abundantly available to optimize these parameters [1-4]. A drawback in the use of post-additions is the additional labor and costs that this fine tuning entails, as well as the environment contamination. [8,9]A typical wood adhesive contains 1 to 15 wt % synthetic additives (e.g. emulsifiers, protective colloids, anti-foam, etc) with respect to the amount of polymer [10]. These additives can be replaced with granular starch, with or without derivatization, where the latter is dissolved and modified just before, or during, the actual polymerization [11,12]. However, the dissolution of granular starch does not only involve a considerable peak viscosity but there is a risk of partial dissolved starch granules as well. Polymerization procedures can be designed to overcome these problems, but there are also starch treatments that can avoid these risks a priori. Extrusion is an example of an energy efficient starch modification which results in a starch product without the presence of starch granules and a viscosity peak during dissolution. Moreover, the amount of water needed during modification is much less than its enzymatic conversion counterpart, for example. Extrusion based processes are therefore frequently favored from a green chemistry and engineering point of view [13,14]. However, the level of modification introduced by extrusion is usually not enough to prevent the amylose part from retrogradation. Therefore, the use of waxy starch is recommended when starch is used without considerable (bio)chemical treatments. The waxy potato (WP) starch is preferred because this type of starch does not only contain the highest amount of phosphate groups, which provides additional stabilization after dissolution, but the granules disintegrate easier after gelatinization as well [15,16].Latexes stabilized with enzymatically converted WP starch fall in the viscosity range of 700 to 2300 mPa·s while the use of extruded regular potato (RP) and WP starch, with and without an octenyl succinylation, results in latexes with viscosities ranging from 1 000 (without derivatization) and beyond 10 000 mPa·s (with derivatization) [17,18]. Both types of latexes are potential wood adhesives and not yet evaluated in literature as such. Modification by extrusion is more energy efficient than enzymatic conversion and the corresponding products are therefore selected for this investigation. The latex stabilized with extruded WP starch
AbstractModified starch can be used as protective colloid in free radical polymerization of polyvinyl acetate. The resulting latexes are potential alternatives for commercial available wood adhesives stabilized with synthetic additives. The wood bonding strength (DIN EN 204 D2) was determined and showed that latexes based on extruded octenyl succinylated (waxy) potato starch are suitable wood adhesives. The wood bonding strengths of the latexes are 2 to 3 times larger than the target 8 MPa value . The use of extruded (octenyl succinylated) waxy potato starch as protective colloid offers flexibility in latex viscosity (1 000 to beyond 10 000 mPa·s) with only limited impact on the wood bonding strength. Mowilith DHS S1 appears to be a proper benchmark for the starch stabilized latexes but this product has a considerably higher latex viscosity (50 000 mPa·s) and wood bonding strength (33 MPa). However, there are indications that the starch based latexes can be tuned to match the commercial benchmark more closely by modifications in the polymerization process.
IntroductionCurrent available synthetic wood adhesives emerged at the beginning of the 20th century and became good alternatives for natural ones (animal, casein, starch, cellulose, etc), of which animal glue being first reported as far as the ancient Egyptian pharaohs times [1,2]. This transition from natural raw materials to synthetic varieties started with the development of phenolic and urea-formaldehyde adhesives (1920) and was followed by the discovery of polychloroprene adhesives after a few decades (1940). Vinyl polymer based emulsion (i.e. latexes) adhesives were invented in the fifties together those based on polyurethane. Another important category in adhesives is a hot melt and this type of product finds its origin in the sixties [1-4].Wood is a heterogeneous material, mainly consisting of fibres aligned in one direction. The material is sensitive for changes in climate conditions, in particular as function of temperature and humidity [1,2]. Climate differences between the manufacturing site and customer countries are therefore usually kept small in order to minimize unnecessary deformation of the final product (e.g. cupboard) during use. The adhesive bonding of wood involves mechanical interlocking (i.e. filling of irregular voids), physical interaction and chemical bonding. Among all possible interaction forces between wood and adhesive, physical ones (e.g. hydrogen bonds and Van der Waals forces) are frequently mentioned as the most relevant ones [1-4]. A durable structural bond requires penetration of the adhesive between several (two to six) layers of fibres and penetration of the fibre cells on a molecular scale [1-4]. Mechanical interlocking is also crucial in the process because the adhesive must also penetrate beyond the damaged wood fibres and fill up voids between the two wood surfaces. Wood adhesives require therefore a good interaction with the substrate (adhesion) and need to have a high internal strength (cohesion) as well [1-4].A standard test (DIN EN 204) has been designed for the evaluation of wood adhesives for non-structural application and divides the adhesives into durability classes D1 to D4 [5]. The division is based on the dry and wet strengths of bond-lines measured under specified conditions after various conditioning treatments. The adhesives that meet this standard are suitable for the use in furniture and interior structures (e.g. panels, doors, windows, stairs
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
102 103
etc…) made of wood or derived timber products. Polyvinyl acetate (pVAc) based adhesives are usually classified as D1 (for use in dry and temperate climates including indoor furniture) or D2 (for use in relative damp indoor, e.g. kitchen furniture). However, some special formulations can even meet D3 (application in damp indoor or outdoor environment) and D4 (water resistant) [5-6]. This set of characteristics and wide applicability spectrum renders these adhesives particularly attractive for further developments.Spray guns, rollers and trowels are frequently used in woodworking and each application technique requires a different level of viscosity. Viscosities in the range of 300 -1 000 mPa·s are suitable for spray application whilst rollers require values in the range of several thousands mPa·s [1,2]. A target value of at least 20 000 mPa·s is required if a trowel is used for application [1,2]. Shear thinning behaviour with a reversible nature (pseudoplasticity or thixotropy) is desired because this improves the workability significantly [7]. This type of adhesive cannot only be added to a substrate more easily but the chance of dripping is reduced as well [7]. Other important application related parameters are pot life, open time, wet tack, curing time and bond strength [1,2]. Additives (plasticizers, fillers, solvents, thickeners and tackifiers) are abundantly available to optimize these parameters [1-4]. A drawback in the use of post-additions is the additional labor and costs that this fine tuning entails, as well as the environment contamination. [8,9]A typical wood adhesive contains 1 to 15 wt % synthetic additives (e.g. emulsifiers, protective colloids, anti-foam, etc) with respect to the amount of polymer [10]. These additives can be replaced with granular starch, with or without derivatization, where the latter is dissolved and modified just before, or during, the actual polymerization [11,12]. However, the dissolution of granular starch does not only involve a considerable peak viscosity but there is a risk of partial dissolved starch granules as well. Polymerization procedures can be designed to overcome these problems, but there are also starch treatments that can avoid these risks a priori. Extrusion is an example of an energy efficient starch modification which results in a starch product without the presence of starch granules and a viscosity peak during dissolution. Moreover, the amount of water needed during modification is much less than its enzymatic conversion counterpart, for example. Extrusion based processes are therefore frequently favored from a green chemistry and engineering point of view [13,14]. However, the level of modification introduced by extrusion is usually not enough to prevent the amylose part from retrogradation. Therefore, the use of waxy starch is recommended when starch is used without considerable (bio)chemical treatments. The waxy potato (WP) starch is preferred because this type of starch does not only contain the highest amount of phosphate groups, which provides additional stabilization after dissolution, but the granules disintegrate easier after gelatinization as well [15,16].Latexes stabilized with enzymatically converted WP starch fall in the viscosity range of 700 to 2300 mPa·s while the use of extruded regular potato (RP) and WP starch, with and without an octenyl succinylation, results in latexes with viscosities ranging from 1 000 (without derivatization) and beyond 10 000 mPa·s (with derivatization) [17,18]. Both types of latexes are potential wood adhesives and not yet evaluated in literature as such. Modification by extrusion is more energy efficient than enzymatic conversion and the corresponding products are therefore selected for this investigation. The latex stabilized with extruded WP starch
AbstractModified starch can be used as protective colloid in free radical polymerization of polyvinyl acetate. The resulting latexes are potential alternatives for commercial available wood adhesives stabilized with synthetic additives. The wood bonding strength (DIN EN 204 D2) was determined and showed that latexes based on extruded octenyl succinylated (waxy) potato starch are suitable wood adhesives. The wood bonding strengths of the latexes are 2 to 3 times larger than the target 8 MPa value . The use of extruded (octenyl succinylated) waxy potato starch as protective colloid offers flexibility in latex viscosity (1 000 to beyond 10 000 mPa·s) with only limited impact on the wood bonding strength. Mowilith DHS S1 appears to be a proper benchmark for the starch stabilized latexes but this product has a considerably higher latex viscosity (50 000 mPa·s) and wood bonding strength (33 MPa). However, there are indications that the starch based latexes can be tuned to match the commercial benchmark more closely by modifications in the polymerization process.
IntroductionCurrent available synthetic wood adhesives emerged at the beginning of the 20th century and became good alternatives for natural ones (animal, casein, starch, cellulose, etc), of which animal glue being first reported as far as the ancient Egyptian pharaohs times [1,2]. This transition from natural raw materials to synthetic varieties started with the development of phenolic and urea-formaldehyde adhesives (1920) and was followed by the discovery of polychloroprene adhesives after a few decades (1940). Vinyl polymer based emulsion (i.e. latexes) adhesives were invented in the fifties together those based on polyurethane. Another important category in adhesives is a hot melt and this type of product finds its origin in the sixties [1-4].Wood is a heterogeneous material, mainly consisting of fibres aligned in one direction. The material is sensitive for changes in climate conditions, in particular as function of temperature and humidity [1,2]. Climate differences between the manufacturing site and customer countries are therefore usually kept small in order to minimize unnecessary deformation of the final product (e.g. cupboard) during use. The adhesive bonding of wood involves mechanical interlocking (i.e. filling of irregular voids), physical interaction and chemical bonding. Among all possible interaction forces between wood and adhesive, physical ones (e.g. hydrogen bonds and Van der Waals forces) are frequently mentioned as the most relevant ones [1-4]. A durable structural bond requires penetration of the adhesive between several (two to six) layers of fibres and penetration of the fibre cells on a molecular scale [1-4]. Mechanical interlocking is also crucial in the process because the adhesive must also penetrate beyond the damaged wood fibres and fill up voids between the two wood surfaces. Wood adhesives require therefore a good interaction with the substrate (adhesion) and need to have a high internal strength (cohesion) as well [1-4].A standard test (DIN EN 204) has been designed for the evaluation of wood adhesives for non-structural application and divides the adhesives into durability classes D1 to D4 [5]. The division is based on the dry and wet strengths of bond-lines measured under specified conditions after various conditioning treatments. The adhesives that meet this standard are suitable for the use in furniture and interior structures (e.g. panels, doors, windows, stairs
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
104 105
of 180 g adhesive per m2 (contact area 625mm2). A climate room was used to maintain the required 23°C and 50% relative humidity during the storage of the samples for 7 days. The samples were dried for 7 days (23°C and 50% relative humidity), soaked in water (3 hours; 23°C) and dried again at 23°C and 50% relative humidity for 7 days before the bonding strength was measured. An Instron 4301 (IX) was used to determine the bonding strength at 50 mm/min. 7 wood pieces were used for each latex.
ResultsThe wood bonding strength of the nine selected starch stabilized latexes were determined in 7 fold. Only averages and two times the standard deviation (σ) are displayed in Table 1 for brevity. The latex viscosity, thermal transitions and the volume mean diameter (VMD) of the particle size distribution (PSD) are displayed in this table as well.
Table 1: Latex characteristics: Viscosity, thermal transitions, VMD and wood bonding strength. Code Latex characteristics Wood test
Viscosity(mPa·s)
Tg,onset (°C)
Tg,inflection (°C)
Tg,endset(°C)
DTg(°C)
VMD(μm)
Average(MPa)
2*s(MPa)
WO2a 9950 8.5 10.6 13.4 4.9 4.2 20.9 12.5WO2b 10253 9.7 11.6 14.4 4.7 4.4 20.8 6.0WO2c 10796 9.2 11.5 13.9 4.7 4.8 20.9 5.9RO2a 2950 9.3 10.9 12.8 3.5 5.4 23.2 2.9RO2b 3000 8.5 11.2 14.0 5.5 5.5 21.5 9.2RO2c 2700 8.8 11.4 13.6 4.8 5.2 21.5 2.3WO0a 1240 10.4 12.8 15.2 4.8 1.7 18.9 6.2WO0b 1300 10.7 13.2 15.1 4.5 1.4 18.5 3.5WO0c 1500 10.3 13.3 15.5 5.2 1.2 17.3 6.3
All average bonding strengths exceed the DIN EN 204 D2 criterion of 8 MPa and the average values of WO2a-c, RO2a-c and WO0a-c are 20.8, 22.0 and 18.3 MPa respectively. The difference between WO2a-c and RO2a-c was small and probably even negligible from an application point of view. The use OSA starch leads to a pronounced better wood bonding strength which might be correlated to the VMD, Tg, latex viscosity or combinations thereof. In this respect, WO0 shows a pronounced difference with respect to RO2 and WO2. Additional research is required to explain the origin of this difference in wood bonding strength. Latex viscosity and hydrophobicity of the starch derivative used are probably the two most interesting responses to investigate because they influence the level of penetration into the wood sample and the susceptibility of the adhesive layer to water.The bonding strength as function of the measuring sequence between the batches of WO2 a-c, RO2 a-c and WO0 a-c (samples: a = 1-7, b = 8-14 & c =15-21) and the observed variation occurs in an essentially random way (Figure 1).
(WO0) is selected together with the latex stabilized with the octenyl succinic anhydride (OSA) modified counterpart with a maximum degree of substitution (DSmax) of 0.02 (WO2). This choice was based on earlier studies showing a latex recovery exceeding 98 wt % and generating a final latex viscosity range of 1 000 to 10 000 mPa·s. The RP starch counterpart (RO2) of WO2 is added in order to garner insight into the effect of using this type of starch as raw material as well. Latexes stabilized with WO0, WO2 and RO2 were prepared in triplicate (a-c) and this allows detecting differences between batches and the effect of mixing them. Method DIN EN 204 D2 was selected for investigating the wood bonding characteristics of the three selected latexes. Commercial available pVAc adhesives Mowilith DHS S1 (DHS) and Mowilith LDL 2555 W (LDL) were used as a benchmark.
ExperimentalMaterialsPVAc latexes stabilized with WO0 (Extruded WP), WO2 (Extruded WP, DSmax OSA 0.02) and RO2 (Extruded RP, DSmax OSA 0.02) [18]. Commercial available wood adhesives Mowilith DHS S1 (DHS) and Mowilith LDL 2555 W (LDL) were selected as benchmarks for DIN EN 204 D2 and D3 respectively.
CharacterizationViscosity, pH and dry matter were determined with the help of a Brookfield DV-II+ (20 RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80°C) respectively. Ethanal and residual vinyl acetate in the latexes were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) gas and a flame ionization detector detector. The gas chromatography measurement was performed on water diluted dispersion (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 minutes and the supernatant was mixed 1:1 with 5 mM NaOH. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. The obtained results were corrected for differences in dry matter content of the latexes evaluated. The dry matter content was used to calculate the amount of water in 100 g of latex and the obtained value was multiplied with the determined concentration of acetate, sulfate and thiosulfate ions.Particle size distributions (PSD) were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (range: 0.13-32.5 μm). Fraunhofer theory based calculations are used and the obtained particle size distributions are ISO 13320 compliant. Glass transition temperatures (Tg) were derived from total heat flow and reversing heat flow curves determined with a modulated differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1°C/min; amplitude: 0.5°C; period: 60 s; large volume stainless steel pans; 20-50 mg dispersion). DIN EN 204 D2 procedures were performed with wood samples made of maple, which were cold pressed with 0.7-0.8 MPa for one hour after application
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
104 105
of 180 g adhesive per m2 (contact area 625mm2). A climate room was used to maintain the required 23°C and 50% relative humidity during the storage of the samples for 7 days. The samples were dried for 7 days (23°C and 50% relative humidity), soaked in water (3 hours; 23°C) and dried again at 23°C and 50% relative humidity for 7 days before the bonding strength was measured. An Instron 4301 (IX) was used to determine the bonding strength at 50 mm/min. 7 wood pieces were used for each latex.
ResultsThe wood bonding strength of the nine selected starch stabilized latexes were determined in 7 fold. Only averages and two times the standard deviation (σ) are displayed in Table 1 for brevity. The latex viscosity, thermal transitions and the volume mean diameter (VMD) of the particle size distribution (PSD) are displayed in this table as well.
Table 1: Latex characteristics: Viscosity, thermal transitions, VMD and wood bonding strength. Code Latex characteristics Wood test
Viscosity(mPa·s)
Tg,onset (°C)
Tg,inflection (°C)
Tg,endset(°C)
DTg(°C)
VMD(μm)
Average(MPa)
2*s(MPa)
WO2a 9950 8.5 10.6 13.4 4.9 4.2 20.9 12.5WO2b 10253 9.7 11.6 14.4 4.7 4.4 20.8 6.0WO2c 10796 9.2 11.5 13.9 4.7 4.8 20.9 5.9RO2a 2950 9.3 10.9 12.8 3.5 5.4 23.2 2.9RO2b 3000 8.5 11.2 14.0 5.5 5.5 21.5 9.2RO2c 2700 8.8 11.4 13.6 4.8 5.2 21.5 2.3WO0a 1240 10.4 12.8 15.2 4.8 1.7 18.9 6.2WO0b 1300 10.7 13.2 15.1 4.5 1.4 18.5 3.5WO0c 1500 10.3 13.3 15.5 5.2 1.2 17.3 6.3
All average bonding strengths exceed the DIN EN 204 D2 criterion of 8 MPa and the average values of WO2a-c, RO2a-c and WO0a-c are 20.8, 22.0 and 18.3 MPa respectively. The difference between WO2a-c and RO2a-c was small and probably even negligible from an application point of view. The use OSA starch leads to a pronounced better wood bonding strength which might be correlated to the VMD, Tg, latex viscosity or combinations thereof. In this respect, WO0 shows a pronounced difference with respect to RO2 and WO2. Additional research is required to explain the origin of this difference in wood bonding strength. Latex viscosity and hydrophobicity of the starch derivative used are probably the two most interesting responses to investigate because they influence the level of penetration into the wood sample and the susceptibility of the adhesive layer to water.The bonding strength as function of the measuring sequence between the batches of WO2 a-c, RO2 a-c and WO0 a-c (samples: a = 1-7, b = 8-14 & c =15-21) and the observed variation occurs in an essentially random way (Figure 1).
(WO0) is selected together with the latex stabilized with the octenyl succinic anhydride (OSA) modified counterpart with a maximum degree of substitution (DSmax) of 0.02 (WO2). This choice was based on earlier studies showing a latex recovery exceeding 98 wt % and generating a final latex viscosity range of 1 000 to 10 000 mPa·s. The RP starch counterpart (RO2) of WO2 is added in order to garner insight into the effect of using this type of starch as raw material as well. Latexes stabilized with WO0, WO2 and RO2 were prepared in triplicate (a-c) and this allows detecting differences between batches and the effect of mixing them. Method DIN EN 204 D2 was selected for investigating the wood bonding characteristics of the three selected latexes. Commercial available pVAc adhesives Mowilith DHS S1 (DHS) and Mowilith LDL 2555 W (LDL) were used as a benchmark.
ExperimentalMaterialsPVAc latexes stabilized with WO0 (Extruded WP), WO2 (Extruded WP, DSmax OSA 0.02) and RO2 (Extruded RP, DSmax OSA 0.02) [18]. Commercial available wood adhesives Mowilith DHS S1 (DHS) and Mowilith LDL 2555 W (LDL) were selected as benchmarks for DIN EN 204 D2 and D3 respectively.
CharacterizationViscosity, pH and dry matter were determined with the help of a Brookfield DV-II+ (20 RPM), WTW pH320 and Mettler Toledo PM100/LP16 (80°C) respectively. Ethanal and residual vinyl acetate in the latexes were determined with a Perkin Elmer gas chromatograph equipped with a headspace sampling device, a Poraplot Q fused silica column (25 m x 0.32 mm) gas and a flame ionization detector detector. The gas chromatography measurement was performed on water diluted dispersion (10 wt %). About 2 ml of the diluted dispersion was centrifuged at 13 000 relative centrifugal force for 10 minutes and the supernatant was mixed 1:1 with 5 mM NaOH. This mixture was used to quantify the anion composition with a Dionex DX50 equipped with an ATC-1 ion trap, two Ionpac columns (AS11-2 mm and AG11-2 mm) and an electrochemical detector. The separation of the different anions was achieved with a gradient of sodium hydroxide. The obtained results were corrected for differences in dry matter content of the latexes evaluated. The dry matter content was used to calculate the amount of water in 100 g of latex and the obtained value was multiplied with the determined concentration of acetate, sulfate and thiosulfate ions.Particle size distributions (PSD) were obtained with a Sympatec laser diffractor equipped with a Quixel wet dispenser and a Helos laser diffraction sensor (range: 0.13-32.5 μm). Fraunhofer theory based calculations are used and the obtained particle size distributions are ISO 13320 compliant. Glass transition temperatures (Tg) were derived from total heat flow and reversing heat flow curves determined with a modulated differential scanning calorimeter (mDSC) from TA Instruments (Q1000; 1°C/min; amplitude: 0.5°C; period: 60 s; large volume stainless steel pans; 20-50 mg dispersion). DIN EN 204 D2 procedures were performed with wood samples made of maple, which were cold pressed with 0.7-0.8 MPa for one hour after application
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
106 107
Figure 2: Box plot calculations for sample sets WO2 a-c, RO2 a-c and WO0 a-c. Outliers are represented by an asterix.
Table 2: Product characteristics: Viscosity, thermal transitions, VMD and wood bonding strength. Code Type Latex characteristics Wood test
Viscosity(mPa·s)
Tg,onset(°C)
Tg,inflection(°C)
Tg,endset(°C)
DTg(°C)
VMD(mm)
Average(MPa)
2*s(MPa)
DHS DIN EN204 D2 49600 12.9 15.4 17.5 4.7 3.1 33.3 2.3LDL DIN EN204 D3 12500 4.1 7.4 10.0 5.9 3.1 39.0 6.2WO2 DIN EN204 D2 10380 8.8 11.6 14.6 5.8 4.4 20.4 6.1RO2 DIN EN204 D2 3020 8.8 11.0 13.3 4.5 5.3 24.4 3.9WO0 DIN EN204 D2 1280 10.7 13.3 15.3 4.6 1.5 17.2 6.8
It is important to bear in mind that the Tg is determined on latex without being dried and that it is based on the reversing heat flow curve. The obtained values differ therefore considerably from the usually reported Tg and this is confirmed by the results of the two commercial available products. The determined values of DHS and LDL with our method (Tg,inflection ) are 15.4 and 7.4 °C whilst their reported Tg are 38 and 26 °C, respectively. The results of the applied method appear to coincide more with the minimum film formation temperature of these products, i.e. 14 and 4 °C, respectively. A correlation is plausible because the minimum film formation temperature and the selected Tg,inflection determination both capture phase transitions of hydrated latex particles. However, this assumption could not be verified in literature.DHS exceeds the starch based products in wood bond strength and differs considerably in viscosity and Tg,inflection. LDL only differs considerably in Tg,inflection but its average bonding strength exceeded the one of DHS. The strength of LDL was close to the integrity strength of the wood samples used because two samples exhibited wood failure and the remaining samples displayed considerable interface adhesive-substrate failure (wood fibres tear). The DHS samples also showed some distinct wood fibres tear but less pronounced and accompanied with domains at which adhesive (cohesive) failure occurred. The starch based varieties showed some wood fibres tear also but adhesive failure was dominating. A Box plot
Figure 1: Wood bonding strength according DIN EN 204 D2 for WO2a-c, RO2 a-c, WO0 a-c and D2 criterion. Sample sets 1-7, 8-14 and 9-21 represent batch a, b and c respectively.
However, the Anderson-Darling goodness of fit calculations (Minitab16) reveal p-values of <0.005, 0.027 and 0.011 for WO2 a-c, RO2 a-c and WO0 a-c respectively. This is considerably lower than the commonly used a = 0.05 criterion and the sample sets are therefore not normally distributed [19]. The origin of this non-normal distribution behaviour is probably related to the inevitable heterogeneity of the wood samples used. The fact that the DIN EN 204 determination requires an average of 20 wood samples is also in agreement with this line of thinking. Figure 1 also shows some exceptional deviations from the general trend which might be considered as outliers. Box plot calculations (Minitab 16) show that the lowest value of sample set WO2a-c and RO2 a-c meet the criterion for outliers and can be removed, if needed, from the sample set (Figure 2). The observed behavior can be conveniently compared with that of commercial formulations based on similar components, with the exception of the protective colloid. DHS is a plasticizer-free aqueous latex suitable for adhesion of wood, paper, packaging, floor, wall, ceiling, leather and textiles based on pVAc and protected by polyvinyl alcohol. LDL is also based on pVAc and can be used in the same area of application except for leather and textiles. LDL has an enhanced resistance to moisture and heat due to the presence of N-methylolacrylamide units which introduce a degree of cross-linking in the final adhesive layer [20]. This difference in composition makes this product very interesting to investigate even if it is classifed as a DIN EN204 D3 type of adhesive (contrary to the D2 target here). WO2, RO2 and WO0 were physical mixtures of the three corresponding batches and the mixtures were tested as well as DHS and LDL according DIN EN204 D2 with 7 wood samples each. The results are given in Table 2 together with the latex viscosity, thermal transitions and the VMD of the PSD.
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
106 107
Figure 2: Box plot calculations for sample sets WO2 a-c, RO2 a-c and WO0 a-c. Outliers are represented by an asterix.
Table 2: Product characteristics: Viscosity, thermal transitions, VMD and wood bonding strength. Code Type Latex characteristics Wood test
Viscosity(mPa·s)
Tg,onset(°C)
Tg,inflection(°C)
Tg,endset(°C)
DTg(°C)
VMD(mm)
Average(MPa)
2*s(MPa)
DHS DIN EN204 D2 49600 12.9 15.4 17.5 4.7 3.1 33.3 2.3LDL DIN EN204 D3 12500 4.1 7.4 10.0 5.9 3.1 39.0 6.2WO2 DIN EN204 D2 10380 8.8 11.6 14.6 5.8 4.4 20.4 6.1RO2 DIN EN204 D2 3020 8.8 11.0 13.3 4.5 5.3 24.4 3.9WO0 DIN EN204 D2 1280 10.7 13.3 15.3 4.6 1.5 17.2 6.8
It is important to bear in mind that the Tg is determined on latex without being dried and that it is based on the reversing heat flow curve. The obtained values differ therefore considerably from the usually reported Tg and this is confirmed by the results of the two commercial available products. The determined values of DHS and LDL with our method (Tg,inflection ) are 15.4 and 7.4 °C whilst their reported Tg are 38 and 26 °C, respectively. The results of the applied method appear to coincide more with the minimum film formation temperature of these products, i.e. 14 and 4 °C, respectively. A correlation is plausible because the minimum film formation temperature and the selected Tg,inflection determination both capture phase transitions of hydrated latex particles. However, this assumption could not be verified in literature.DHS exceeds the starch based products in wood bond strength and differs considerably in viscosity and Tg,inflection. LDL only differs considerably in Tg,inflection but its average bonding strength exceeded the one of DHS. The strength of LDL was close to the integrity strength of the wood samples used because two samples exhibited wood failure and the remaining samples displayed considerable interface adhesive-substrate failure (wood fibres tear). The DHS samples also showed some distinct wood fibres tear but less pronounced and accompanied with domains at which adhesive (cohesive) failure occurred. The starch based varieties showed some wood fibres tear also but adhesive failure was dominating. A Box plot
Figure 1: Wood bonding strength according DIN EN 204 D2 for WO2a-c, RO2 a-c, WO0 a-c and D2 criterion. Sample sets 1-7, 8-14 and 9-21 represent batch a, b and c respectively.
However, the Anderson-Darling goodness of fit calculations (Minitab16) reveal p-values of <0.005, 0.027 and 0.011 for WO2 a-c, RO2 a-c and WO0 a-c respectively. This is considerably lower than the commonly used a = 0.05 criterion and the sample sets are therefore not normally distributed [19]. The origin of this non-normal distribution behaviour is probably related to the inevitable heterogeneity of the wood samples used. The fact that the DIN EN 204 determination requires an average of 20 wood samples is also in agreement with this line of thinking. Figure 1 also shows some exceptional deviations from the general trend which might be considered as outliers. Box plot calculations (Minitab 16) show that the lowest value of sample set WO2a-c and RO2 a-c meet the criterion for outliers and can be removed, if needed, from the sample set (Figure 2). The observed behavior can be conveniently compared with that of commercial formulations based on similar components, with the exception of the protective colloid. DHS is a plasticizer-free aqueous latex suitable for adhesion of wood, paper, packaging, floor, wall, ceiling, leather and textiles based on pVAc and protected by polyvinyl alcohol. LDL is also based on pVAc and can be used in the same area of application except for leather and textiles. LDL has an enhanced resistance to moisture and heat due to the presence of N-methylolacrylamide units which introduce a degree of cross-linking in the final adhesive layer [20]. This difference in composition makes this product very interesting to investigate even if it is classifed as a DIN EN204 D3 type of adhesive (contrary to the D2 target here). WO2, RO2 and WO0 were physical mixtures of the three corresponding batches and the mixtures were tested as well as DHS and LDL according DIN EN204 D2 with 7 wood samples each. The results are given in Table 2 together with the latex viscosity, thermal transitions and the VMD of the PSD.
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
108 109
As a consequence it is difficult to speculate that the degree of polydispersity alone explains the observed differences in bonding strength. Similarly, the difference in viscosity (and thus in level of penetration of the adhesive into the wood) cannot solely explain the observed differences between DHS, WO2 and RO2 either. A proportional relation between latex viscosity and bonding strength could be an explanation for DHS (~50 000 mPa·s; 33 MPa) with respect to either WO2 (~10 000 mPa·s; 20 MPa) or RO2 (~3 000 mPa·s; 24 MPa). However, the results of WO2 and RO2 display an inverse proportional relationship between both responses. On the other hand, the measured Tg,inflection indicated a structural difference in pVAc between DHS (15 °C; 33 MPa) and both starch based varieties (12-11 °C; 20-24 MPa). This seems to correlate with the bonding strength. A lower Tg is associated with a difference in composition or branching of the polymer and indications of the occurrence of the latter were already found at the polymerization stage. The reaction temperature of WO0 a-c during preparation, for example, was considerably lower than its WO2 a-c and RO2 a-c counterparts and resulted in a pronounced difference in Tg,inflection [18]. The level of branching was inversely proportional to the Tg and DHS should have a lower degree of branching from this line of reasoning. A lower degree of branching should increase the cohesive strength of the adhesive and this was actually the case. The wood bond strengths of WO2 and RO2 were limited due to adhesive (cohesive) failure whilst the failure of DHS was dominated by wood fibre tear (a sign of a higher cohesive strength). Specific details about the polymerization procedure of DHS and LDL are not known and several product responses (i.e. dry matter, hydrogen ion, ethanal, VAM, acetate, sulfate and thiosulfate content) are therefore determined. The results are displayed in Table 3 together with the results of the three selected starch stabilized counterparts.
Table 3: Product composition: dry matter, pH, ethanal, VAM, acetate, sulfate and thiosulfate.
Code TypeDry
matter(mg/g)
Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol) pH Ethanal
(mg/g)VAM
(mg/g)
DHS DIN EN204 D2 492 139 52.4 n.d. 4.6 n.d. 0.2LDL DIN EN204 D3 494 199 5.9 n.d. 3.1 n.d. 0.1WO2 DIN EN204 D2 521 324 30.0 2.42 5.3 2.0 1.5RO2 DIN EN204 D2 514 330 30.3 2.46 5.3 3.1 1.6WO0 DIN EN204 D2 514 279 31.2 2.74 5.3 3.2 1.0
n.d. = not detected
WO2, RO2 and WO0 contained similar amounts of free acetate and the level found was significantly higher than in DHS and LDL. Acetate is partly generated during the anion determination because the sample treatment involves a pH at which saponification of pVAc can occur rapidly. The water phase will still contain some water soluble pVAc after the centrifugation step and this will be totally converted to acetate after the NaOH addition in the sample treatment. The actual amount of water soluble pVAc in the latexes is not known and additional research is therefore required if the actual origin of the acetate ions needs to be known. WO2, RO2 and WO0 were prepared in the same way according the responses sulfate, thiosulfate and pH whilst DHS and LDL show pronounced differences between each
shows that only the set of WO2 contains an outlier and that all test results exceed the DIN EN204 D2 criterion of 8 MPa (Figure 3).
Figure 3: Box plot calculations for sample sets DHS, LDL, WO2, RO2 and WO0. Outliers are represented by an asterix.
There was a distinct difference between the two benchmark latexes and the starch stabilized counterparts. The superior bonding characteristic of LDL finds its origin in the fact that it is able to crosslink whereas the other four products lack this feature. The differences found between DHS and the starch based counterparts were probably related to viscosity level, PSD, nature of protective colloid used, pVAc structure and combinations thereof. The deviating behaviour of WO0 with respect to WO2 and RO2 can probably be attributed to its lower viscosity and smaller VMD. These two features make a higher degree of penetration in the wood structure possible and this process leaves less adhesive available for the actual bond formation between the two substrates. The observed differences in particle size between DHS, WO2 and RO2 were not only modest in VMD but in their PSD also (Figure 4).
Figure 4: PSD’s of WO2, RO2 and DHS. Error bars represents 2 times the σ of a measurement in five fold (DHS only)
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
108 109
As a consequence it is difficult to speculate that the degree of polydispersity alone explains the observed differences in bonding strength. Similarly, the difference in viscosity (and thus in level of penetration of the adhesive into the wood) cannot solely explain the observed differences between DHS, WO2 and RO2 either. A proportional relation between latex viscosity and bonding strength could be an explanation for DHS (~50 000 mPa·s; 33 MPa) with respect to either WO2 (~10 000 mPa·s; 20 MPa) or RO2 (~3 000 mPa·s; 24 MPa). However, the results of WO2 and RO2 display an inverse proportional relationship between both responses. On the other hand, the measured Tg,inflection indicated a structural difference in pVAc between DHS (15 °C; 33 MPa) and both starch based varieties (12-11 °C; 20-24 MPa). This seems to correlate with the bonding strength. A lower Tg is associated with a difference in composition or branching of the polymer and indications of the occurrence of the latter were already found at the polymerization stage. The reaction temperature of WO0 a-c during preparation, for example, was considerably lower than its WO2 a-c and RO2 a-c counterparts and resulted in a pronounced difference in Tg,inflection [18]. The level of branching was inversely proportional to the Tg and DHS should have a lower degree of branching from this line of reasoning. A lower degree of branching should increase the cohesive strength of the adhesive and this was actually the case. The wood bond strengths of WO2 and RO2 were limited due to adhesive (cohesive) failure whilst the failure of DHS was dominated by wood fibre tear (a sign of a higher cohesive strength). Specific details about the polymerization procedure of DHS and LDL are not known and several product responses (i.e. dry matter, hydrogen ion, ethanal, VAM, acetate, sulfate and thiosulfate content) are therefore determined. The results are displayed in Table 3 together with the results of the three selected starch stabilized counterparts.
Table 3: Product composition: dry matter, pH, ethanal, VAM, acetate, sulfate and thiosulfate.
Code TypeDry
matter(mg/g)
Acetate(mmol)
Sulfate(mmol)
Thiosulfate(mmol) pH Ethanal
(mg/g)VAM
(mg/g)
DHS DIN EN204 D2 492 139 52.4 n.d. 4.6 n.d. 0.2LDL DIN EN204 D3 494 199 5.9 n.d. 3.1 n.d. 0.1WO2 DIN EN204 D2 521 324 30.0 2.42 5.3 2.0 1.5RO2 DIN EN204 D2 514 330 30.3 2.46 5.3 3.1 1.6WO0 DIN EN204 D2 514 279 31.2 2.74 5.3 3.2 1.0
n.d. = not detected
WO2, RO2 and WO0 contained similar amounts of free acetate and the level found was significantly higher than in DHS and LDL. Acetate is partly generated during the anion determination because the sample treatment involves a pH at which saponification of pVAc can occur rapidly. The water phase will still contain some water soluble pVAc after the centrifugation step and this will be totally converted to acetate after the NaOH addition in the sample treatment. The actual amount of water soluble pVAc in the latexes is not known and additional research is therefore required if the actual origin of the acetate ions needs to be known. WO2, RO2 and WO0 were prepared in the same way according the responses sulfate, thiosulfate and pH whilst DHS and LDL show pronounced differences between each
shows that only the set of WO2 contains an outlier and that all test results exceed the DIN EN204 D2 criterion of 8 MPa (Figure 3).
Figure 3: Box plot calculations for sample sets DHS, LDL, WO2, RO2 and WO0. Outliers are represented by an asterix.
There was a distinct difference between the two benchmark latexes and the starch stabilized counterparts. The superior bonding characteristic of LDL finds its origin in the fact that it is able to crosslink whereas the other four products lack this feature. The differences found between DHS and the starch based counterparts were probably related to viscosity level, PSD, nature of protective colloid used, pVAc structure and combinations thereof. The deviating behaviour of WO0 with respect to WO2 and RO2 can probably be attributed to its lower viscosity and smaller VMD. These two features make a higher degree of penetration in the wood structure possible and this process leaves less adhesive available for the actual bond formation between the two substrates. The observed differences in particle size between DHS, WO2 and RO2 were not only modest in VMD but in their PSD also (Figure 4).
Figure 4: PSD’s of WO2, RO2 and DHS. Error bars represents 2 times the σ of a measurement in five fold (DHS only)
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
110 111
AbbreviationsDIN EN 204: Wood adhesive test for non-structural application.
- D2 : For use in relative damp indoor (e.g. kitchen furniture).- D3 : Application in damp indoor or outdoor environment.
pVAc : Polyvinyl acetate.WP : Waxy potato starch.RP : Regular potato starch.OSA : Octenyl succinic anhydride.DSmax : Maximum degree of substitution based on the amount of reagent added.WO0 : Extruded WP starch.WO2 : Extruded WP starch with OSA a derivatization of 0.02 (DSmax).RO2 : Extruded RP starch with OSA a derivatization of 0.02 (DSmax).XXXa-c : Triplicate.DHS : Mowilith DHS S1LDL : Mowilith LDL 2555 WPSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulated Differential Calorimeter.σ : Standard deviation.VMD : Volume Mean Diameter.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
other and the three starch stabilized latexes. A persulfate initiation system is still plausible according the sulfate content of DHS (entry 1) but the absence of thiosulfate is indicative for the use of an initiation system in combination with a different reducing agent (or none at all). The use of persulfate in the preparation process of LDL is highly unlikely based on the sulfate content of this latex (entry 2). The preparation process of DHS and LDL appeared to be much more efficient with respect to residual ethanal and VAM than their starch based counterparts. However, these two commercial products might have had a post-treatment to achieve this level of VAM conversion but there is no evidence found that such a post-treatment is actually applied. The amount of volatile organic components in the latex is bound to stringent regulations and there are different techniques (e.g. vacuum or steam stripping) available to meet these regulations properly [21]. The presence of volatile organic components in the latex might have a side effect due to the fact it can be stored in the particles and act as a plasticizer [22]. However, the amounts of volatile organic components found in the starch based latexes are very low and a considerable impact on the Tg, for example, is therefore not expected.From the considerations above, it is clear that a full and comprehensive comparison of the adhesive performance between the prepared formulations and the two commercial ones is still possible at a qualitative level. A more detailed (i.e. molecular) comparison would be needed in order to fully relate the observed differences with the structure of the corresponding latexes.
ConclusionsThe evaluated potato starch stabilized latexes can be marked as DIN EN204 D2 wood adhesives because they have wood bonding strengths in the range of 17 – 24 MPa whilst only 8 MPa is required for this classification. An octenyl succinylation of the starch used did not only increase the latex viscosity but improved the wood bonding strength as well. The latex viscosity of octenyl succinylated waxy potato based products was 3 to 4 times higher than its regular potato counterpart.Mowilith DHS S1 appears to be a proper benchmark for the starch derivative protected latexes because a product composition evaluation points to a main difference in the use of protective colloid only. The wood bonding strength of this product was 33 MPa and its viscosity of 50 000 mPa·s was also considerably higher than the 17-24 MPa and the 1 280 – 12 500 mPa·s for the evaluated starch based counterparts. However, there were indications of structural differences in the polyvinyl acetate present in Mowilith DHS S1 and the starch stabilized varieties which might be reduced significantly by a change in polymerization conditions. A change in level of octenyl succinate derivatization or degree of extrusion induced degradation might also offer opportunities to reduce, or even close, the gap between the starch protected latexes and formulations based on Mowilith DHS S1.
AcknowledgementsThis investigation was sponsored by Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives
110 111
AbbreviationsDIN EN 204: Wood adhesive test for non-structural application.
- D2 : For use in relative damp indoor (e.g. kitchen furniture).- D3 : Application in damp indoor or outdoor environment.
pVAc : Polyvinyl acetate.WP : Waxy potato starch.RP : Regular potato starch.OSA : Octenyl succinic anhydride.DSmax : Maximum degree of substitution based on the amount of reagent added.WO0 : Extruded WP starch.WO2 : Extruded WP starch with OSA a derivatization of 0.02 (DSmax).RO2 : Extruded RP starch with OSA a derivatization of 0.02 (DSmax).XXXa-c : Triplicate.DHS : Mowilith DHS S1LDL : Mowilith LDL 2555 WPSD : Particle size distribution.Tg : Glass transition temperature.mDSC : Modulated Differential Calorimeter.σ : Standard deviation.VMD : Volume Mean Diameter.Tg,onset : Onset point based glass transition temperature.Tg,inflection : Inflection point based glass transition temperature.Tg,endset : Endset point based glass transition temperature.DTg : Tg,endset – Tg,onset.
other and the three starch stabilized latexes. A persulfate initiation system is still plausible according the sulfate content of DHS (entry 1) but the absence of thiosulfate is indicative for the use of an initiation system in combination with a different reducing agent (or none at all). The use of persulfate in the preparation process of LDL is highly unlikely based on the sulfate content of this latex (entry 2). The preparation process of DHS and LDL appeared to be much more efficient with respect to residual ethanal and VAM than their starch based counterparts. However, these two commercial products might have had a post-treatment to achieve this level of VAM conversion but there is no evidence found that such a post-treatment is actually applied. The amount of volatile organic components in the latex is bound to stringent regulations and there are different techniques (e.g. vacuum or steam stripping) available to meet these regulations properly [21]. The presence of volatile organic components in the latex might have a side effect due to the fact it can be stored in the particles and act as a plasticizer [22]. However, the amounts of volatile organic components found in the starch based latexes are very low and a considerable impact on the Tg, for example, is therefore not expected.From the considerations above, it is clear that a full and comprehensive comparison of the adhesive performance between the prepared formulations and the two commercial ones is still possible at a qualitative level. A more detailed (i.e. molecular) comparison would be needed in order to fully relate the observed differences with the structure of the corresponding latexes.
ConclusionsThe evaluated potato starch stabilized latexes can be marked as DIN EN204 D2 wood adhesives because they have wood bonding strengths in the range of 17 – 24 MPa whilst only 8 MPa is required for this classification. An octenyl succinylation of the starch used did not only increase the latex viscosity but improved the wood bonding strength as well. The latex viscosity of octenyl succinylated waxy potato based products was 3 to 4 times higher than its regular potato counterpart.Mowilith DHS S1 appears to be a proper benchmark for the starch derivative protected latexes because a product composition evaluation points to a main difference in the use of protective colloid only. The wood bonding strength of this product was 33 MPa and its viscosity of 50 000 mPa·s was also considerably higher than the 17-24 MPa and the 1 280 – 12 500 mPa·s for the evaluated starch based counterparts. However, there were indications of structural differences in the polyvinyl acetate present in Mowilith DHS S1 and the starch stabilized varieties which might be reduced significantly by a change in polymerization conditions. A change in level of octenyl succinate derivatization or degree of extrusion induced degradation might also offer opportunities to reduce, or even close, the gap between the starch protected latexes and formulations based on Mowilith DHS S1.
AcknowledgementsThis investigation was sponsored by Samenwerkingsverband Noord-Nederland (SNN) and the Province of Groningen, ordinance Transitie II and Pieken.
112
CHAPTER 7Technological assessment
References[1] P. Cognard, Adhesive bonding of wood and wood based products Part 1 till 4, (http://www.omnexus4adhesives.
com; 27-03-2013) [2] P. Cognard, Adhesives and sealants: General knowledge, Application Technique, New curing techniques,
Elsevier Ltd, Oxford, 2006 [3] I. Skeist, Handbook of adhesives, 3rd edition, Chapman & Hall, New York, 1989.[4] W.H. Cubberly, R. Bakerjian, Tool and Manufacturing Engineers Handbook, Volume 5, Society of
Manufacturing Engineers, Michigan, 1989[5] DIN EN 204, Classification of thermoplastic wood adhesives for non-structural applications, 2001, (http://
webstore.ansi.org/RecordDetail.aspx?sku=DIN+EN+204%3A2001; 01-02-2013)[6] J. Lu, A.J. Easteal, N.R. Edmonds, Crosslinkable poly(vinyl acetate) emulsions for wood adhesive, Pigment
& Resin Technology, 40 (3) (2011) 161-168[7] H.A. Barnes, Thixotropy: A review, J. Non-Newtonian Fluid Mech., 70 (1997) 1-33[8] Henkel, Water-based Adhesives as low migration products, ( http://www.henkel.com/com/content_
data/20120503_Drupa_Adhesin_Low_Migration_eng.pdf; 7-12-2012)[9] P. Jones, Dispersion polymers: Technology & applications, 22 (11) (2011), 1[10] T. Zanetta, F. Chiozza, C. Rei, Wood adhesive compositions (2013) WO2013057214A1 [11] C.P. Iovine, Y.J. Shih, P.R. Mudge, P.T. Trzasko, Hydrophobically modified starch stabilized vinyl ester
polymer emulsions (1986) EP0223145[12] S.M. Hurley, F.L. Toss, P.E. Sandvick, S.E. Danley, Starch degradation/graft polymerization composition,
process, and uses thereof, (2000) US6090884[13] ACS Green Chemistry Institute, The twelve principles of green engineering, www.acs.org/greenchemistry
(18-03-2013)[14] ACS Green Chemistry Institute, The twelve principles of green chemistry, www.acs.org/greenchemistry
(18-03-2013)[15] J.J.M. Swinkels, Composition and properties of commercial native starches, Starch, 37 (1) (1987) 1-5[16] A.M. Hermansson, K. Svegmark, Developments in the understanding of starch functionality, Trends in food
science & technology, 7 (1996) 345[17] K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified waxy potato starch
stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics, To be published.
[18] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and waxy potato starch, To be published.
[19] RAC Start sheet, Anderson-Darling: Goodness of Fit test for small sample assumptions 10 (2003) 5 (http://src.alionscience.com/pdf/A_DTest.pdf; 27-03-2013)
[20] C. Balzarek, M. Jakob, Cross-Linkable Polymer Dispersions, Method for the Production Thereof and Use Thereof (2010) US0179272
[21] J. Bohling, P.F. Doll, D.L. Frattarelli, K.R. Manna, A.M. Maurice, Low odor compositions and low odor coating compositions (2011) US0152406.
[22] D.M.C. Heymans, M.F. Daniel, Glass transition and film formation of veova/vinyl acetate latices; Role of water and co-solvents, Polymers for advanced technologies 6 (1995) 291.
112
CHAPTER 7Technological assessment
References[1] P. Cognard, Adhesive bonding of wood and wood based products Part 1 till 4, (http://www.omnexus4adhesives.
com; 27-03-2013) [2] P. Cognard, Adhesives and sealants: General knowledge, Application Technique, New curing techniques,
Elsevier Ltd, Oxford, 2006 [3] I. Skeist, Handbook of adhesives, 3rd edition, Chapman & Hall, New York, 1989.[4] W.H. Cubberly, R. Bakerjian, Tool and Manufacturing Engineers Handbook, Volume 5, Society of
Manufacturing Engineers, Michigan, 1989[5] DIN EN 204, Classification of thermoplastic wood adhesives for non-structural applications, 2001, (http://
webstore.ansi.org/RecordDetail.aspx?sku=DIN+EN+204%3A2001; 01-02-2013)[6] J. Lu, A.J. Easteal, N.R. Edmonds, Crosslinkable poly(vinyl acetate) emulsions for wood adhesive, Pigment
& Resin Technology, 40 (3) (2011) 161-168[7] H.A. Barnes, Thixotropy: A review, J. Non-Newtonian Fluid Mech., 70 (1997) 1-33[8] Henkel, Water-based Adhesives as low migration products, ( http://www.henkel.com/com/content_
data/20120503_Drupa_Adhesin_Low_Migration_eng.pdf; 7-12-2012)[9] P. Jones, Dispersion polymers: Technology & applications, 22 (11) (2011), 1[10] T. Zanetta, F. Chiozza, C. Rei, Wood adhesive compositions (2013) WO2013057214A1 [11] C.P. Iovine, Y.J. Shih, P.R. Mudge, P.T. Trzasko, Hydrophobically modified starch stabilized vinyl ester
polymer emulsions (1986) EP0223145[12] S.M. Hurley, F.L. Toss, P.E. Sandvick, S.E. Danley, Starch degradation/graft polymerization composition,
process, and uses thereof, (2000) US6090884[13] ACS Green Chemistry Institute, The twelve principles of green engineering, www.acs.org/greenchemistry
(18-03-2013)[14] ACS Green Chemistry Institute, The twelve principles of green chemistry, www.acs.org/greenchemistry
(18-03-2013)[15] J.J.M. Swinkels, Composition and properties of commercial native starches, Starch, 37 (1) (1987) 1-5[16] A.M. Hermansson, K. Svegmark, Developments in the understanding of starch functionality, Trends in food
science & technology, 7 (1996) 345[17] K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified waxy potato starch
stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics, To be published.
[18] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and waxy potato starch, To be published.
[19] RAC Start sheet, Anderson-Darling: Goodness of Fit test for small sample assumptions 10 (2003) 5 (http://src.alionscience.com/pdf/A_DTest.pdf; 27-03-2013)
[20] C. Balzarek, M. Jakob, Cross-Linkable Polymer Dispersions, Method for the Production Thereof and Use Thereof (2010) US0179272
[21] J. Bohling, P.F. Doll, D.L. Frattarelli, K.R. Manna, A.M. Maurice, Low odor compositions and low odor coating compositions (2011) US0152406.
[22] D.M.C. Heymans, M.F. Daniel, Glass transition and film formation of veova/vinyl acetate latices; Role of water and co-solvents, Polymers for advanced technologies 6 (1995) 291.
Technological assessment
114 115
because it is a waste of energy and it lowers the temperature of the reaction mixture. Moreover, the polymerization process might become suboptimal at reflux conditions or even terminate in case the free radicals are (also) generated by thermal dissociation. A reflux induced drop in reaction temperature frequently occurred during attempts to convert available polymerization procedures into a reference procedure based on the thermal dissociation of persulfate [1,3,4,15-20]. As a consequence, the desired polymerization procedure in this study had to be designed from scratch [21]. The defined method is suitable for generating latexes with viscosities ranging from water thin to a paste (Figure 1).
Figure 1: The consistency of the latexes prepared during the investigation of this thesis. From left to right: latexes with a low, moderate and high level of viscosity.
The reactor configuration and polymerization procedures used in this study were not yet fully optimized. It is recommended to use fundamental mathematical models in case a thorough optimization is desired [22]. These kinds of models are not only very powerful tools for optimization but are also suitable for scale-up, process control, monitoring, operator training and often allow for a better understanding of underlying mechanisms.
Final productCommercially available homopolymers of polyvinyl acetate are usually blended with other materials for an optimal price/performance ratio. In this respect, the high wood bonding strength (EN204;D2) of Mowilith DHS S1 (33 MPa) offers more possibilities for blending with (cheap) materials than the evaluated potato starch stabilized latexes (17 - 24 MPa) [23]. However, the bonding strength of the latter might be improved considerably by making a switch from a homopolymer to a polyvinyl acetate based copolymer. Monomers with a hydrophobic character (e.g. butyl acrylate) or cross-linking abilities (e.g. low formaldehyde releasing N-methylol acrylamide) might be interesting options to investigate if the bonding strength of latex needs to be changed [12].
A switch from a vinyl acetate homopolymer to a copolymer will not only influence the bonding strength of the latex but the glass transition temperature (Tg) will be changed as well. Monomer types that lower the Tg of the latex are preferred because that might reduce the need of plasticizers in the final product. Plasticizers are frequently added to vinyl acetate
Starch as protective colloidThe use of native starch as protective colloid is not only interesting from a cost-price perspective but also according the principles of green chemistry – these principles favour ingredients with minimal pre-treatments. Native starch can be used to stabilize synthetic latexes as long as it is (partly) dissolved before the actual free radical polymerization is initiated [1]. Unfortunately, this approach has not only benefits. Firstly, the dissolution of starch granules takes time and this reduces the manufacturing throughput if this pre-treatment is executed in the polymerization reactor. In the second place, native starch dissolves with a considerable peak in viscosity and additional measures, with respect to proper mixing, might be needed to avoid unacceptable heterogeneities during processing. Thirdly, the starch molecules need to be degraded, either mechanical or (bio)chemical), before the polymerization can be started. And finally, native starch is hydrophilic and starch with a slight hydrophobic nature might be required for some important latex characteristics (e.g. rheology and particle size distribution) [2-4].
Polymerization procedures can be tuned to ensure proper dissolution and degradation of native starch, but there are also starch pre-treatments which can incorporate the desired starch characteristics a priori. Extrusion is an example of an energy efficient starch modification which results in products without the presence of granules and a peak viscosity during dissolution. When compared to other modification strategies, several advantages can be outlined. For instance, the amount of water needed during this kind of starch modification is less than its enzymatic counterpart. As a consequence, enzymatic treatments require more energy during dissolution of starch granules and drying of the final product. For this reason, extrusion is an interesting starch pre-treatment from a sustainability point of view [5-8].
Hydrophobic (e.g. octenyl succinylated) starch has emulsifying properties and can therefore be used to change the particle formation process during polymerization and/or the rheology of the final latex [2-4,9-12]. The octenyl succinylation step of the in thesis investigated extruded starches was performed in suspension (~39 wt % in water) and the obtained product needed to be (partly) dried before extrusion. As a result, it is interesting to investigate if it is possible to execute this derivatization in line at semi-dry conditions (< 30 wt % water) just before the extrusion step [13]. This preparation process is considerably less energy demanding than its suspension counterpart for less water needs to be removed during drying. Moreover, there are other ways to prepare hydrophobic starches in an efficient way. The superheated steam driven acylation of starch is an interesting one because fatty acids, which are more environmental benign than octenyl succinic acid for example, can be used [14].
Free radical reactor configuration and polymerization procedureIt is common practice in the latex industry to execute vinyl acetate polymerizations above the boiling point of the azeotrope water and vinyl acetate monomer (66 °C). This precondition renders the desire to start with a formulation with modified starch as only additive quite a challenging one. Indeed, the absence of emulsifiers does not only affect the particle formation process, but increases the concentration of vinyl acetate monomer in the water phase as well [10,12]. The latter increases the risk of excessive refluxing and this need to be avoided
Technological assessment
114 115
because it is a waste of energy and it lowers the temperature of the reaction mixture. Moreover, the polymerization process might become suboptimal at reflux conditions or even terminate in case the free radicals are (also) generated by thermal dissociation. A reflux induced drop in reaction temperature frequently occurred during attempts to convert available polymerization procedures into a reference procedure based on the thermal dissociation of persulfate [1,3,4,15-20]. As a consequence, the desired polymerization procedure in this study had to be designed from scratch [21]. The defined method is suitable for generating latexes with viscosities ranging from water thin to a paste (Figure 1).
Figure 1: The consistency of the latexes prepared during the investigation of this thesis. From left to right: latexes with a low, moderate and high level of viscosity.
The reactor configuration and polymerization procedures used in this study were not yet fully optimized. It is recommended to use fundamental mathematical models in case a thorough optimization is desired [22]. These kinds of models are not only very powerful tools for optimization but are also suitable for scale-up, process control, monitoring, operator training and often allow for a better understanding of underlying mechanisms.
Final productCommercially available homopolymers of polyvinyl acetate are usually blended with other materials for an optimal price/performance ratio. In this respect, the high wood bonding strength (EN204;D2) of Mowilith DHS S1 (33 MPa) offers more possibilities for blending with (cheap) materials than the evaluated potato starch stabilized latexes (17 - 24 MPa) [23]. However, the bonding strength of the latter might be improved considerably by making a switch from a homopolymer to a polyvinyl acetate based copolymer. Monomers with a hydrophobic character (e.g. butyl acrylate) or cross-linking abilities (e.g. low formaldehyde releasing N-methylol acrylamide) might be interesting options to investigate if the bonding strength of latex needs to be changed [12].
A switch from a vinyl acetate homopolymer to a copolymer will not only influence the bonding strength of the latex but the glass transition temperature (Tg) will be changed as well. Monomer types that lower the Tg of the latex are preferred because that might reduce the need of plasticizers in the final product. Plasticizers are frequently added to vinyl acetate
Starch as protective colloidThe use of native starch as protective colloid is not only interesting from a cost-price perspective but also according the principles of green chemistry – these principles favour ingredients with minimal pre-treatments. Native starch can be used to stabilize synthetic latexes as long as it is (partly) dissolved before the actual free radical polymerization is initiated [1]. Unfortunately, this approach has not only benefits. Firstly, the dissolution of starch granules takes time and this reduces the manufacturing throughput if this pre-treatment is executed in the polymerization reactor. In the second place, native starch dissolves with a considerable peak in viscosity and additional measures, with respect to proper mixing, might be needed to avoid unacceptable heterogeneities during processing. Thirdly, the starch molecules need to be degraded, either mechanical or (bio)chemical), before the polymerization can be started. And finally, native starch is hydrophilic and starch with a slight hydrophobic nature might be required for some important latex characteristics (e.g. rheology and particle size distribution) [2-4].
Polymerization procedures can be tuned to ensure proper dissolution and degradation of native starch, but there are also starch pre-treatments which can incorporate the desired starch characteristics a priori. Extrusion is an example of an energy efficient starch modification which results in products without the presence of granules and a peak viscosity during dissolution. When compared to other modification strategies, several advantages can be outlined. For instance, the amount of water needed during this kind of starch modification is less than its enzymatic counterpart. As a consequence, enzymatic treatments require more energy during dissolution of starch granules and drying of the final product. For this reason, extrusion is an interesting starch pre-treatment from a sustainability point of view [5-8].
Hydrophobic (e.g. octenyl succinylated) starch has emulsifying properties and can therefore be used to change the particle formation process during polymerization and/or the rheology of the final latex [2-4,9-12]. The octenyl succinylation step of the in thesis investigated extruded starches was performed in suspension (~39 wt % in water) and the obtained product needed to be (partly) dried before extrusion. As a result, it is interesting to investigate if it is possible to execute this derivatization in line at semi-dry conditions (< 30 wt % water) just before the extrusion step [13]. This preparation process is considerably less energy demanding than its suspension counterpart for less water needs to be removed during drying. Moreover, there are other ways to prepare hydrophobic starches in an efficient way. The superheated steam driven acylation of starch is an interesting one because fatty acids, which are more environmental benign than octenyl succinic acid for example, can be used [14].
Free radical reactor configuration and polymerization procedureIt is common practice in the latex industry to execute vinyl acetate polymerizations above the boiling point of the azeotrope water and vinyl acetate monomer (66 °C). This precondition renders the desire to start with a formulation with modified starch as only additive quite a challenging one. Indeed, the absence of emulsifiers does not only affect the particle formation process, but increases the concentration of vinyl acetate monomer in the water phase as well [10,12]. The latter increases the risk of excessive refluxing and this need to be avoided
Technological assessment
116 117
starch during the initial stage of the polymerization process might be a viable route if the desired properties of the latex require the presence of a hydrophobic starch derivative during processing. The mechanism how to achieve this type of grafting on polysaccharides is frequently described in literature but not yet in a way that can be easily translated to an industrial setting. For example, the industry needs guidance in controlling the level of grafting in a large semi-batch (or small continuous) reactor whilst most investigations provide information about grafting on a small scale in batch mode. As a result, the latex industry is waiting for research especially designed to make the available fundamental knowledge about grafting more accessible to them.
homopolymers for optimal performance but they tend to migrate out of the adhesive layer after application. The need of plasticizer in the final product might become even superfluous if ethylene, or di-alkyl maleate, based copolymers are used [12].
The rheology of synthetic latex is complex because it is a liquid multiphase system with incompatible components. As a result, the microstructure of these systems can consist of droplets in a matrix, elongated fibrils or a co-continuous structure [24]. The level of viscosity that latex generates is therefore not a very good variable to describe the differences and similarities between different varieties prepared. Moreover, waxy potato starch tends to associate with hydrophobic chains and this interaction will influence the level of viscosity of the corresponding latex as well [2,25]. Large Amplitude Oscillatory Shear (LAOS) characterization, often referred in the literature as Fourier Transform Rheology (FTR), might be helpful in comparing the different available latexes to each other [24]. LAOS possesses a high sensitivity in the characterization of the morphology, thus allowing evaluation of properties that might otherwise be missed with traditional linear methodologies
Starch can also be used as rheology modifier in those cases that the latex is prepared with starch as main stabilizing agent. Modifiers are usually added to the product in a separate blending silo after the polymerization is finished. No relevant studies were found during a quick screening of open literature, dealing with such two-step envisaged procedure and their actual need in practice. In addition, partial replacement of polyvinyl acetate in the polymerization recipe by a cheap rheology modifier might result in latexes similar to those prepared by the traditional two step approach. The accompanying loss in value of the latex prepared might be (partly) compensated by a shorter polymerization procedure, an easier preparation process and less handling losses. However, a polymerization reactor is more difficult to handle than a blending silo and more expensive in use as well.
A molar replacement of vinyl acetate monomer by other water soluble monomers should not be very problematic in the described preparation procedures of this study. Latexes based on hydrophobic monomers can also be prepared as long as there is sufficient affinity between the monomer and waxy potato starch fragments (or a hydrophobic starch derivative is used). This approach might unlock other application areas (e.g. paint, coating or ink) were (waxy) potato starch can be used as main stabilizing agent in free radical based polymerizations as well. Furthermore, the technology to prepare vinyl acetate and ethylene monomers from a renewable feedstock is available and the corresponding copolymer is already presented as the ideal binder for environmental friendly paints. It would be interesting to investigate the actual added value of these green polymers with respect to their conventional counterparts and make an estimate of the turning point at which these green polymers become commercial viable if they are stabilized with starch [26,27].
It is an improvement, from the point of view of sustainability, if oil-based and migration-sensitive ingredients are replaced by counterparts made of starch with only a slight chemical derivatization. However, getting a product safety legislation approval of a starch derivative is more difficult than starch without any chemical derivatization. Grafting of monomer onto
Technological assessment
116 117
starch during the initial stage of the polymerization process might be a viable route if the desired properties of the latex require the presence of a hydrophobic starch derivative during processing. The mechanism how to achieve this type of grafting on polysaccharides is frequently described in literature but not yet in a way that can be easily translated to an industrial setting. For example, the industry needs guidance in controlling the level of grafting in a large semi-batch (or small continuous) reactor whilst most investigations provide information about grafting on a small scale in batch mode. As a result, the latex industry is waiting for research especially designed to make the available fundamental knowledge about grafting more accessible to them.
homopolymers for optimal performance but they tend to migrate out of the adhesive layer after application. The need of plasticizer in the final product might become even superfluous if ethylene, or di-alkyl maleate, based copolymers are used [12].
The rheology of synthetic latex is complex because it is a liquid multiphase system with incompatible components. As a result, the microstructure of these systems can consist of droplets in a matrix, elongated fibrils or a co-continuous structure [24]. The level of viscosity that latex generates is therefore not a very good variable to describe the differences and similarities between different varieties prepared. Moreover, waxy potato starch tends to associate with hydrophobic chains and this interaction will influence the level of viscosity of the corresponding latex as well [2,25]. Large Amplitude Oscillatory Shear (LAOS) characterization, often referred in the literature as Fourier Transform Rheology (FTR), might be helpful in comparing the different available latexes to each other [24]. LAOS possesses a high sensitivity in the characterization of the morphology, thus allowing evaluation of properties that might otherwise be missed with traditional linear methodologies
Starch can also be used as rheology modifier in those cases that the latex is prepared with starch as main stabilizing agent. Modifiers are usually added to the product in a separate blending silo after the polymerization is finished. No relevant studies were found during a quick screening of open literature, dealing with such two-step envisaged procedure and their actual need in practice. In addition, partial replacement of polyvinyl acetate in the polymerization recipe by a cheap rheology modifier might result in latexes similar to those prepared by the traditional two step approach. The accompanying loss in value of the latex prepared might be (partly) compensated by a shorter polymerization procedure, an easier preparation process and less handling losses. However, a polymerization reactor is more difficult to handle than a blending silo and more expensive in use as well.
A molar replacement of vinyl acetate monomer by other water soluble monomers should not be very problematic in the described preparation procedures of this study. Latexes based on hydrophobic monomers can also be prepared as long as there is sufficient affinity between the monomer and waxy potato starch fragments (or a hydrophobic starch derivative is used). This approach might unlock other application areas (e.g. paint, coating or ink) were (waxy) potato starch can be used as main stabilizing agent in free radical based polymerizations as well. Furthermore, the technology to prepare vinyl acetate and ethylene monomers from a renewable feedstock is available and the corresponding copolymer is already presented as the ideal binder for environmental friendly paints. It would be interesting to investigate the actual added value of these green polymers with respect to their conventional counterparts and make an estimate of the turning point at which these green polymers become commercial viable if they are stabilized with starch [26,27].
It is an improvement, from the point of view of sustainability, if oil-based and migration-sensitive ingredients are replaced by counterparts made of starch with only a slight chemical derivatization. However, getting a product safety legislation approval of a starch derivative is more difficult than starch without any chemical derivatization. Grafting of monomer onto
Technological assessment
118 119
process for their manufacture (1997) EP799837[26] N. Alperowicz, WACKER eyes green polymers production, Chemical week 4 (5) (2010) Vol. 172
Issue 8, 23.[27] W. Huster, Vinyl acetate-ethylene copolymer dispersions ideal binders for environmental friendly
paints, WACKER Polymers (http://www.sesam-uae.com/sustainablematerials/presentations/02_wacker_huster.pdf; 22-05-2011)
References[1] S.M. Hurley, F.L. Toss, P.E. Sandvick, S.E. Danley, Starch degradation/graft polymerization composition,
process, and uses thereof, (2000) US6090884[2] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Extruded octenyl
succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and amylopectin potato starch, To be published.
[3] R.L. Billmers, R. Farwaha, G.S. Yearwood, L. Phan, Thixotropic paint compositions containing hydrophobic starch derivatives (1999) US6001927
[4] C.P. Iovine, Y.J. Shih, P.R. Mudge, P.T. Trzasko, Hydrophobically modified starch stabilized vinyl ester polymer emulsions (1986) EP0223145
[5] ACS Green Chemistry Institute, The twelve principles of green engineering, (www.acs.org;18-03-2013)[6] ACS Green Chemistry Institute, The twelve principles of green chemistry, (www.acs.org; 18-03-2013)[7] R.M. Van den Einde, A.J. Van der Goot, R.M. Boom, Understanding molecular weight reduction of starch
during heating-shearing processes, Journal of food science, 68 (8) (2003) 2396-2404[8] C. Bramsiepe, S. Sievers, T. Seifert, G.D. Stefanidis, D.G. Vlachos, H. Schnitzer, B. Muster, C. Brunner,
J.P.M. Sanders, M.E. Bruins, G. Schembecker, Low-cost small scale processing technologies for production applications in various environments-mass produced factories, Chem. Eng. Process 51 (2012) 32-52.
[9] J.M. Ponce-Ortega, M.M. Al-Thubaiti, M.M. El-Halwagi, Process intensification: New understanding and systematic approach, Chem. Eng. Process 53 (2012) 63-75.
[10] H. B. Yamak, Emulsion Polymerization: Effects of Polymerization Variables on the Properties of Vinyl Acetate Based Emulsion Polymers, Polymer Science, Dr. Faris Yılmaz (Ed.), Intech, 2013 (http://www.intechopen.com/books/polymer-science; 20-12-2013)
[11] H.J. De Vries, C. Semeijn, P.L. Buwalda, Emulsifier (2005) EP1743693[12] H. Warson, C.A. Finch, Application of synthetic resin lattices: Fundamental chemistry of latices and
applications in adhesives, (2001) John Wiley and Sons Ltd. [13] E. Hadderingh, A.A.M. Maas, R.P.W. Kesselmans, H.C. Hiemstra, Cross-linking of starch (2002)
EP1390412.[14] I.K. Haaksman, J.W. Timmermans, T.M. Slaghek, Acylation of carbohydrates (2009) EP 2062923.[15] P. F.T. Lambrechts, J.H.R. van der Meeren, Method for preparing a water-containing vinyl acetate polymer
dispersion, dispersion thus prepared and protective colloid used thereby (1979) EP 0021542[16] L. Kovats, Starch derivative protective colloids in emulsion polymer systems (1971) US 3769248[17] A.E. Cohen, P.L. Gordon, Water paint comprising pigment and aqueous emulsion of polyvinyl acetate and
waxy maize starch (1956) US2914495[18] M. Pfalz, M.W. Marnik, D. Gruel, Emulsion polymerization method (1999) EP1232189. [19] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum
adhesive based on a filled polymer dispersion, WO2006094594A1. [20] A.P. Kightlinger, E.L. Speakman, G.T. Van Duzee, Stable. Liquid, Amylopectin starch graft copolymer
compositions (1980) GB2075526.[21] K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman, H.J.
Heeres, Modified amylopectin potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects. To be published
[22] P.A. Mueller, J.R. Richards, J. P. Congalidis, Polymerization Reactor Modeling in Industry. Macromolecular Reaction Engineering, 5: 261–277
[23] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives, To be published.
[24] G. Nikolic, Fourier Transforms - New Analytical Approaches and FTIR Strategies, InTech (2011) 285-302.[25] H. Huizenga, H.G. Alrich, I.L. De Groot, Aqueous compositions comprising amylopectin-potato starch and
Technological assessment
118 119
process for their manufacture (1997) EP799837[26] N. Alperowicz, WACKER eyes green polymers production, Chemical week 4 (5) (2010) Vol. 172
Issue 8, 23.[27] W. Huster, Vinyl acetate-ethylene copolymer dispersions ideal binders for environmental friendly
paints, WACKER Polymers (http://www.sesam-uae.com/sustainablematerials/presentations/02_wacker_huster.pdf; 22-05-2011)
References[1] S.M. Hurley, F.L. Toss, P.E. Sandvick, S.E. Danley, Starch degradation/graft polymerization composition,
process, and uses thereof, (2000) US6090884[2] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Extruded octenyl
succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and amylopectin potato starch, To be published.
[3] R.L. Billmers, R. Farwaha, G.S. Yearwood, L. Phan, Thixotropic paint compositions containing hydrophobic starch derivatives (1999) US6001927
[4] C.P. Iovine, Y.J. Shih, P.R. Mudge, P.T. Trzasko, Hydrophobically modified starch stabilized vinyl ester polymer emulsions (1986) EP0223145
[5] ACS Green Chemistry Institute, The twelve principles of green engineering, (www.acs.org;18-03-2013)[6] ACS Green Chemistry Institute, The twelve principles of green chemistry, (www.acs.org; 18-03-2013)[7] R.M. Van den Einde, A.J. Van der Goot, R.M. Boom, Understanding molecular weight reduction of starch
during heating-shearing processes, Journal of food science, 68 (8) (2003) 2396-2404[8] C. Bramsiepe, S. Sievers, T. Seifert, G.D. Stefanidis, D.G. Vlachos, H. Schnitzer, B. Muster, C. Brunner,
J.P.M. Sanders, M.E. Bruins, G. Schembecker, Low-cost small scale processing technologies for production applications in various environments-mass produced factories, Chem. Eng. Process 51 (2012) 32-52.
[9] J.M. Ponce-Ortega, M.M. Al-Thubaiti, M.M. El-Halwagi, Process intensification: New understanding and systematic approach, Chem. Eng. Process 53 (2012) 63-75.
[10] H. B. Yamak, Emulsion Polymerization: Effects of Polymerization Variables on the Properties of Vinyl Acetate Based Emulsion Polymers, Polymer Science, Dr. Faris Yılmaz (Ed.), Intech, 2013 (http://www.intechopen.com/books/polymer-science; 20-12-2013)
[11] H.J. De Vries, C. Semeijn, P.L. Buwalda, Emulsifier (2005) EP1743693[12] H. Warson, C.A. Finch, Application of synthetic resin lattices: Fundamental chemistry of latices and
applications in adhesives, (2001) John Wiley and Sons Ltd. [13] E. Hadderingh, A.A.M. Maas, R.P.W. Kesselmans, H.C. Hiemstra, Cross-linking of starch (2002)
EP1390412.[14] I.K. Haaksman, J.W. Timmermans, T.M. Slaghek, Acylation of carbohydrates (2009) EP 2062923.[15] P. F.T. Lambrechts, J.H.R. van der Meeren, Method for preparing a water-containing vinyl acetate polymer
dispersion, dispersion thus prepared and protective colloid used thereby (1979) EP 0021542[16] L. Kovats, Starch derivative protective colloids in emulsion polymer systems (1971) US 3769248[17] A.E. Cohen, P.L. Gordon, Water paint comprising pigment and aqueous emulsion of polyvinyl acetate and
waxy maize starch (1956) US2914495[18] M. Pfalz, M.W. Marnik, D. Gruel, Emulsion polymerization method (1999) EP1232189. [19] O. Sommer, H. Buxhoffer, N. De Calmes, R. Gossen, S. Kotthoff, H.J. Wolter, E. Abrahams-Meyer, Gum
adhesive based on a filled polymer dispersion, WO2006094594A1. [20] A.P. Kightlinger, E.L. Speakman, G.T. Van Duzee, Stable. Liquid, Amylopectin starch graft copolymer
compositions (1980) GB2075526.[21] K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman, H.J.
Heeres, Modified amylopectin potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects. To be published
[22] P.A. Mueller, J.R. Richards, J. P. Congalidis, Polymerization Reactor Modeling in Industry. Macromolecular Reaction Engineering, 5: 261–277
[23] K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives, To be published.
[24] G. Nikolic, Fourier Transforms - New Analytical Approaches and FTIR Strategies, InTech (2011) 285-302.[25] H. Huizenga, H.G. Alrich, I.L. De Groot, Aqueous compositions comprising amylopectin-potato starch and
120
APPENDICESSummary SamenvattingDankwoordPublications
120
APPENDICESSummary SamenvattingDankwoordPublications
Summary
122 123
available on a commercial scale and also contains some unique properties with respect to viscosity, associative behaviour with hydrophobic moieties and stability after dissolution. Waxy potato starch products generate higher viscosities and stabilities than their regular potato starch counterparts, whereas the latter exceeds products made of regular cereal or tapioca starch in viscosity and stability significantly.
Synthetic latexes are usually prepared in large polymerization reactors (i.e. several cubic meters) because this is a commercial accepted way towards producing low cost products. The polymerization procedures used usually are already completely optimized and the use of cheaper ingredients appears to be the easiest way to reduce the costs even more. However, a change of ingredients will not only impact the properties of the final latex, but the characteristics of the reaction mixture as well. Therefore, the price reduction obtained by the use of cheaper ingredients can, if not addressed properly, result in inferior products and/or more expensive production processes. As a result, the use of other ingredients in latexes often requires a thorough re-evaluation of the remaining part of the formulation and the corresponding polymerization process. Several attempts to stabilize current commercial available latexes with modified starch instead of synthetic additives, failed to produce a commercial acceptable product in the past. Sufficient leads can be found in academic and industrial related archives to solve the encountered problem. However, most information is rather obsolete because western society changed dramatically in the past decades. Firstly, health and environmental related issues are very important these days. In the second place, commonly used raw materials of the past will become scarce in the near future. Thirdly, energy and labour costs are a major part of the cost price. And finally, customers are nowadays used to products with a very high level of quality. Consequently, the manufacturing of starch stabilized latexes is in need of thorough redesign before the corresponding latexes can seriously compete with some of the currently available products with respect to quality and price. It is also important to evaluate all steps of the life cycle of the product and focus on those applications in which the (chemical) modification of the starch can be kept to a minimum. This is not only beneficial from a cost price point of view, but also for its level of sustainability and biodegradability.
There is a preference for applications that require good adherence between the latex particles and cellulose based materials. Addition of starch to latex might result in synergistic effects in adhesion, because starch has a high affinity with cellulose. This line of thinking makes latex based paints, inks, coatings and adhesives for wood and paper based materials the most interesting areas to investigate. The best starting point for this investigation appears to be wood glues with polyvinyl acetate as main ingredient. On the one hand, starch is a potential alternative for the protective colloid (i.e. polyvinyl alcohol) that is frequently used in this type of glue. On the other hand, the compatibility between starch and polyvinyl acetate is good. To conclude, the wood glue market is potentially large enough to justify a thorough investigation in replacing the polyvinyl alcohol part of the glue by starch.
Several free radical polymerization procedures for preparing polyvinyl acetate latexes were found after a screening of accessible information. These collected preparation procedures
Product development is on the one hand more and more directed by environment and human related issues but on the other hand, choices that result in a lower cost price are still favoured. Nevertheless, the worldwide demand for sustainable products is steadily increasing despite the fact that they are usually slightly more expensive than their conventional counterparts. This direction is often supported by governmental legislation and the actual price of a sustainable product is often artificially reduced by a great variety of subsidies. However, it is still quite a challenge to improve the level of sustainability of existing products despite the available social and financial incentives. Production plants are designed and organised in a specific way for an optimal production of conventional but not sustainable products. Besides, customers are used to certain, sometimes secondary, product characteristics. For example, in some cases specific post-additions are required for optimal performance and the alternative latex might need different post-additions to get the same properties as the original product. Furthermore, the life time of products is currently rapidly decreasing and therefore, a shift from large non-flexible production processes towards smaller flexible counterparts is desired. As a result, adaptations in all parts of the value chain are required to successfully introduce new, more sustainable products.
Synthetic latexes are basic ingredients for commercially available adhesives, coatings, inks and paints and worldwide over 10 million metric tons are annually produced. The latexes are usually prepared by free radical polymerizations executed in semi-batch reactors. This way of production requires intensive mixing because reagents are added during processing and the polymerization reaction involved is highly exothermic. Control of the reaction temperature and the degree of mixing during the process is crucial and the level of control increases with the surface area to volume ratio of the reactor used. As a consequence, a reduction in the size of the reactor coincides with an improved level of reaction control. However, smaller reactors make the preparation process less economically interesting in case of (semi-) batch based processes. Switching to small continuous reactors is attractive from a commercial point of view, but this direction requires a considerable change in plant organisation and layout. The fact that the semi-batch process is still favoured in the emulsion polymerization industry over their continuous counterparts, who have been developed over a decade ago, suggests that this switch is commercially not attractive enough yet.
Starch-stabilized synthetic latexes have been commercially available for approximately 60 years but the application of synthetic additives as stabilizers is currently dominating the biggest markets (i.e. paint, coating and adhesives). Starch is a renewable raw material that is considerably more water soluble than for example cellulose. This makes starch on the one hand better biodegradable and on the other considerably more susceptible for modification and derivatization in this environmental benign solvent. Starch can originate from different plant sources of which maize (corn), wheat, cassava (tapioca) and potato are most important for application on an industrial scale. Each type of starch has its own specific set of benefits and drawbacks that make them more or less interesting for certain applications. Potato starch is renowned for its low lipid and protein content and the amylopectin (also known as waxy) variety has even a higher purity due to its very low amylose content. The latter is only recently
Summary
122 123
available on a commercial scale and also contains some unique properties with respect to viscosity, associative behaviour with hydrophobic moieties and stability after dissolution. Waxy potato starch products generate higher viscosities and stabilities than their regular potato starch counterparts, whereas the latter exceeds products made of regular cereal or tapioca starch in viscosity and stability significantly.
Synthetic latexes are usually prepared in large polymerization reactors (i.e. several cubic meters) because this is a commercial accepted way towards producing low cost products. The polymerization procedures used usually are already completely optimized and the use of cheaper ingredients appears to be the easiest way to reduce the costs even more. However, a change of ingredients will not only impact the properties of the final latex, but the characteristics of the reaction mixture as well. Therefore, the price reduction obtained by the use of cheaper ingredients can, if not addressed properly, result in inferior products and/or more expensive production processes. As a result, the use of other ingredients in latexes often requires a thorough re-evaluation of the remaining part of the formulation and the corresponding polymerization process. Several attempts to stabilize current commercial available latexes with modified starch instead of synthetic additives, failed to produce a commercial acceptable product in the past. Sufficient leads can be found in academic and industrial related archives to solve the encountered problem. However, most information is rather obsolete because western society changed dramatically in the past decades. Firstly, health and environmental related issues are very important these days. In the second place, commonly used raw materials of the past will become scarce in the near future. Thirdly, energy and labour costs are a major part of the cost price. And finally, customers are nowadays used to products with a very high level of quality. Consequently, the manufacturing of starch stabilized latexes is in need of thorough redesign before the corresponding latexes can seriously compete with some of the currently available products with respect to quality and price. It is also important to evaluate all steps of the life cycle of the product and focus on those applications in which the (chemical) modification of the starch can be kept to a minimum. This is not only beneficial from a cost price point of view, but also for its level of sustainability and biodegradability.
There is a preference for applications that require good adherence between the latex particles and cellulose based materials. Addition of starch to latex might result in synergistic effects in adhesion, because starch has a high affinity with cellulose. This line of thinking makes latex based paints, inks, coatings and adhesives for wood and paper based materials the most interesting areas to investigate. The best starting point for this investigation appears to be wood glues with polyvinyl acetate as main ingredient. On the one hand, starch is a potential alternative for the protective colloid (i.e. polyvinyl alcohol) that is frequently used in this type of glue. On the other hand, the compatibility between starch and polyvinyl acetate is good. To conclude, the wood glue market is potentially large enough to justify a thorough investigation in replacing the polyvinyl alcohol part of the glue by starch.
Several free radical polymerization procedures for preparing polyvinyl acetate latexes were found after a screening of accessible information. These collected preparation procedures
Product development is on the one hand more and more directed by environment and human related issues but on the other hand, choices that result in a lower cost price are still favoured. Nevertheless, the worldwide demand for sustainable products is steadily increasing despite the fact that they are usually slightly more expensive than their conventional counterparts. This direction is often supported by governmental legislation and the actual price of a sustainable product is often artificially reduced by a great variety of subsidies. However, it is still quite a challenge to improve the level of sustainability of existing products despite the available social and financial incentives. Production plants are designed and organised in a specific way for an optimal production of conventional but not sustainable products. Besides, customers are used to certain, sometimes secondary, product characteristics. For example, in some cases specific post-additions are required for optimal performance and the alternative latex might need different post-additions to get the same properties as the original product. Furthermore, the life time of products is currently rapidly decreasing and therefore, a shift from large non-flexible production processes towards smaller flexible counterparts is desired. As a result, adaptations in all parts of the value chain are required to successfully introduce new, more sustainable products.
Synthetic latexes are basic ingredients for commercially available adhesives, coatings, inks and paints and worldwide over 10 million metric tons are annually produced. The latexes are usually prepared by free radical polymerizations executed in semi-batch reactors. This way of production requires intensive mixing because reagents are added during processing and the polymerization reaction involved is highly exothermic. Control of the reaction temperature and the degree of mixing during the process is crucial and the level of control increases with the surface area to volume ratio of the reactor used. As a consequence, a reduction in the size of the reactor coincides with an improved level of reaction control. However, smaller reactors make the preparation process less economically interesting in case of (semi-) batch based processes. Switching to small continuous reactors is attractive from a commercial point of view, but this direction requires a considerable change in plant organisation and layout. The fact that the semi-batch process is still favoured in the emulsion polymerization industry over their continuous counterparts, who have been developed over a decade ago, suggests that this switch is commercially not attractive enough yet.
Starch-stabilized synthetic latexes have been commercially available for approximately 60 years but the application of synthetic additives as stabilizers is currently dominating the biggest markets (i.e. paint, coating and adhesives). Starch is a renewable raw material that is considerably more water soluble than for example cellulose. This makes starch on the one hand better biodegradable and on the other considerably more susceptible for modification and derivatization in this environmental benign solvent. Starch can originate from different plant sources of which maize (corn), wheat, cassava (tapioca) and potato are most important for application on an industrial scale. Each type of starch has its own specific set of benefits and drawbacks that make them more or less interesting for certain applications. Potato starch is renowned for its low lipid and protein content and the amylopectin (also known as waxy) variety has even a higher purity due to its very low amylose content. The latter is only recently
Summary
124 125
vinyl acetate monomer can be stored in these aggregates. The presence of these aggregates improves the level of control of the polymerization with respect to level of refluxing and particle formation. Octenyl succinylated starch has distinct emulsifier characteristics and the impact of this type of starch on the polymerization process and latex properties is investigated in chapter 5. The starch products were extruded (i.e. modified by mechanical shear) before use in order to introduce a modest level of degradation and this treatment made them cold water soluble as well. The starch fragments present after dissolution are very large with respect to commonly used and recommended starch derivatives, which are usually degraded to the oligomeric level. Latexes stabilized with amylopectin potato starch showed the most remarkable change after an octenyl succinylation. An increase of ~1000 to ~10000 mPa·s was observed for the amylopectin variety with a degree of substitution of 0.02 whilst its regular counterpart only showed an increase of ~1000 to ~3000 mPa·s after introducing the same level of octenyl succinylation.
Application of glue with rollers is a very popular technique in the wood working industry and requires viscosities of at least thousand mPa·s. A number of the (octenyl succinylated) (amylopectin) potato starch stabilized latexes prepared meet this criterion and three of them were checked for wood bonding characteristics in chapter 6. The investigation is based on standard wood adhesive test “EN204 D2” and this test is representative for wood adhesive application in areas with only limited exposure to moist. The three evaluated latexes showed wood bonding strengths in the magnitude of 17 to 24 MPa and this is considerably higher than the 8 MPa criterion of the selected wood bonding test.
were not only checked in practice but also evaluated according to the principles of green chemistry - a guideline to increase the level of sustainability of synthetic products. The obtained knowledge was used to design the polymerization system and procedure described in chapter 2. In this chapter, enzymatic converted amylopectin potato starch (10 wt % on monomer) was used as stabilizer in absence of synthetic additives (i.e. emulsifier, anti-foam and/or protective colloid). The corresponding polymerizations were executed in the presence of inhibitors (i.e. oxygen and hydroquinone) and the influence of three independent variables (i.e. level of agitation, pre-dosage of monomer and pre-dosage initiator/buffer mixture) were evaluated. Heat flow during processing, monomer conversion, product recovery, anion concentration, pH, viscosity, particle size distribution, amount of grafted protective colloid and glass transition temperature were the responses determined. The reproducibility level of the designed preparation procedure appears to be sufficient with respect to the variation observed at considerably different settings. The obtained results reveal a number of leads which can be helpful if these latexes need to be optimized with respect to functionality and level of sustainability.
The pyrodextrination of starch is mainly influenced by heat exposure, level of moisture and hydrochloric acid content. Chapter 3 illustrates the influence of changes in reaction conditions on the properties of potato starch based pyrodextrins and their stabilizing characteristics. Changes in hydrochloric acid content and heat exposure showed a higher impact on the stabilizing properties of the pyrodextrins than the applied differences in level of pre-drying. The pyrodextrins with the highest energy consumption during the pyrodextrination process resulted in latexes with the most desired latex characteristic (i.e. monodisperse particle size distribution). Moreover, the preparations of these latexes were more energy demanding as well, due to a lot of refluxing (i.e. waste of energy) during processing. A considerable reduction in energy demand in the preparation process of this type of latexes seams feasible, nevertheless additional research is needed to determine the optimal process conditions for both pyrodextrination and polymerization.
The degree of sustainability of a given latex can be increased by a reduction in energy consumption during processing, an increase in dry solids content or a combination thereof. Chapter 4 shows the impact of changes in this area for polyvinyl acetate latex stabilized with enzymatic converted potato amylopectin starch (25 wt % on monomer). Reaction temperatures of 75 to 85 °C were applied and the concentration of initiator varied from 2 to 4 wt %. A randomized 22 factorial design augmented with five centre points was selected to investigate the influence of these two factors on several product and process related responses. Considerable differences in reaction conditions were induced but all polymerizations were probably executed at the boiling point of the azeotrope vinyl acetate monomer and water (66 °C). The viscosities of the obtained latexes varied from 700 to 2300 mPa·s and a statistical model was derived in an attempt to correlate the latex viscosity with both processing factors (i.e. reaction temperature and initiator concentration).
Emulsifiers generate hydrophobic aggregates after dissolution in water and excess
Summary
124 125
vinyl acetate monomer can be stored in these aggregates. The presence of these aggregates improves the level of control of the polymerization with respect to level of refluxing and particle formation. Octenyl succinylated starch has distinct emulsifier characteristics and the impact of this type of starch on the polymerization process and latex properties is investigated in chapter 5. The starch products were extruded (i.e. modified by mechanical shear) before use in order to introduce a modest level of degradation and this treatment made them cold water soluble as well. The starch fragments present after dissolution are very large with respect to commonly used and recommended starch derivatives, which are usually degraded to the oligomeric level. Latexes stabilized with amylopectin potato starch showed the most remarkable change after an octenyl succinylation. An increase of ~1000 to ~10000 mPa·s was observed for the amylopectin variety with a degree of substitution of 0.02 whilst its regular counterpart only showed an increase of ~1000 to ~3000 mPa·s after introducing the same level of octenyl succinylation.
Application of glue with rollers is a very popular technique in the wood working industry and requires viscosities of at least thousand mPa·s. A number of the (octenyl succinylated) (amylopectin) potato starch stabilized latexes prepared meet this criterion and three of them were checked for wood bonding characteristics in chapter 6. The investigation is based on standard wood adhesive test “EN204 D2” and this test is representative for wood adhesive application in areas with only limited exposure to moist. The three evaluated latexes showed wood bonding strengths in the magnitude of 17 to 24 MPa and this is considerably higher than the 8 MPa criterion of the selected wood bonding test.
were not only checked in practice but also evaluated according to the principles of green chemistry - a guideline to increase the level of sustainability of synthetic products. The obtained knowledge was used to design the polymerization system and procedure described in chapter 2. In this chapter, enzymatic converted amylopectin potato starch (10 wt % on monomer) was used as stabilizer in absence of synthetic additives (i.e. emulsifier, anti-foam and/or protective colloid). The corresponding polymerizations were executed in the presence of inhibitors (i.e. oxygen and hydroquinone) and the influence of three independent variables (i.e. level of agitation, pre-dosage of monomer and pre-dosage initiator/buffer mixture) were evaluated. Heat flow during processing, monomer conversion, product recovery, anion concentration, pH, viscosity, particle size distribution, amount of grafted protective colloid and glass transition temperature were the responses determined. The reproducibility level of the designed preparation procedure appears to be sufficient with respect to the variation observed at considerably different settings. The obtained results reveal a number of leads which can be helpful if these latexes need to be optimized with respect to functionality and level of sustainability.
The pyrodextrination of starch is mainly influenced by heat exposure, level of moisture and hydrochloric acid content. Chapter 3 illustrates the influence of changes in reaction conditions on the properties of potato starch based pyrodextrins and their stabilizing characteristics. Changes in hydrochloric acid content and heat exposure showed a higher impact on the stabilizing properties of the pyrodextrins than the applied differences in level of pre-drying. The pyrodextrins with the highest energy consumption during the pyrodextrination process resulted in latexes with the most desired latex characteristic (i.e. monodisperse particle size distribution). Moreover, the preparations of these latexes were more energy demanding as well, due to a lot of refluxing (i.e. waste of energy) during processing. A considerable reduction in energy demand in the preparation process of this type of latexes seams feasible, nevertheless additional research is needed to determine the optimal process conditions for both pyrodextrination and polymerization.
The degree of sustainability of a given latex can be increased by a reduction in energy consumption during processing, an increase in dry solids content or a combination thereof. Chapter 4 shows the impact of changes in this area for polyvinyl acetate latex stabilized with enzymatic converted potato amylopectin starch (25 wt % on monomer). Reaction temperatures of 75 to 85 °C were applied and the concentration of initiator varied from 2 to 4 wt %. A randomized 22 factorial design augmented with five centre points was selected to investigate the influence of these two factors on several product and process related responses. Considerable differences in reaction conditions were induced but all polymerizations were probably executed at the boiling point of the azeotrope vinyl acetate monomer and water (66 °C). The viscosities of the obtained latexes varied from 700 to 2300 mPa·s and a statistical model was derived in an attempt to correlate the latex viscosity with both processing factors (i.e. reaction temperature and initiator concentration).
Emulsifiers generate hydrophobic aggregates after dissolution in water and excess
126
APPENDICESSummary SamenvattingDankwoordPublications
126
APPENDICESSummary SamenvattingDankwoordPublications
Samenvatting
128 129
kan afkomstig zijn van verschillende plantaardige bronnen waarvan maïs, tarwe, cassave (tapioca) en aardappel de belangrijkste zijn voor toepassing op industriële schaal. Elk type zetmeel heeft zijn eigen specifieke set van voor- en nadelen waardoor ze meer of minder interessant voor bepaalde toepassingen zijn. Aardappelzetmeel staat bekend om zijn lage vet- en eiwitgehalte en de amylopectine (meestal aangeduid met de term waxy) variant heeft zelfs een nog hogere homogeniteit vanwege het zeer lage gehalte aan amylose. De laatste is sinds kort op commerciële schaal beschikbaar en heeft ook een aantal unieke eigenschappen met betrekking tot viscositeit, associatief gedrag met hydrofobe groepen en stabiliteit na oplossen. Waxy aardappelzetmeel producten genereren een hogere viscositeit en stabiliteit dan de tegenhangers op basis van gewoon aardappelzetmeel en de laatste resulteert weer in aanzienlijk meer viscositeit en stabiliteit dan producten gemaakt van regulier graan of tapioca zetmeel.
Synthetische latexen worden meestal bereid in polymerisatiereactoren van enkele kubieke meters want dit is een commercieel breed geaccepteerde manier om goedkope producten te maken. De bijbehorende polymerisatie procedures zijn vaak al volledig geoptimaliseerd en het gebruik van goedkopere ingrediënten lijkt dan de meest eenvoudige manier om de kosten nog verder verlagen. Echter, een verandering in formulering heeft niet alleen invloed op de eigenschappen van het uiteindelijke product maar ook op de eigenschappen van het reactiemengsel. Het toepassen van goedkopere ingrediënten kan, in het ergste geval, tot inferieure producten en/of duurdere productieprocessen leiden. De impact van de aanpassing, op zowel het resterende deel van de formulering als bijbehorend polymerisatieproces, moet daarom ook eerst goed onderzocht worden. Verschillende proeven om de synthetische hulpstoffen in commerciële latexen te vervangen door gemodificeerd zetmeel zijn in het verleden al uitgevoerd maar het is nog niet gelukt om een commercieel aantrekkelijk alternatief te maken. Er zijn voldoende relevante publicaties beschikbaar (tot wel 60 jaar oud) die als uitgangspunt kunnen dienen om de optredende problemen op te lossen. Alleen is de beschikbare informatie enigszins achterhaald omdat de westerse samenleving drastisch is veranderd in de afgelopen decennia. Ten eerste, gezondheid en milieu gerelateerde onderwerpen zijn tegenwoordig bijvoorbeeld veel belangrijker geworden. Ten tweede, veel in het verleden gebruikte ingrediënten worden steeds schaarser of sommige additieven mogen nu niet meer worden gebruikt. Ten derde, hebben energie- en arbeidskosten nu een veel groter aandeel in de kostprijs dan vroeger. En tot slot, zijn gebruikers nu aan producten met een veel hoger kwaliteitsniveau gewend. Voor de beschikbare zetmeel gebaseerde latexen is dus ook extra onderzoek nodig voordat ze serieus kunnen concurreren met de huidige latexen die gestabiliseerd zijn met synthetische hulpstoffen. Ook is het belangrijk om alle stappen van de levenscyclus van het product te beoordelen en het onderzoek te richten op die toepassingen waarin de (chemische) modificatie van het zetmeel tot een minimum kunnen worden beperkt. Dit is niet alleen gunstig vanuit het oogpunt van kostprijs, maar ook voor het niveau van duurzaamheid en biologische afbreekbaarheid.
Er is een voorkeur voor toepassingen die een goede hechting tussen de latex deeltjes en cellulose gebaseerde materialen vereisen. Het toevoegen van zetmeel aan de betreffende
Productontwikkeling wordt aan de ene kant meer en meer beïnvloed door mens en milieu gerelateerde zaken maar aan de andere kant is er nog steeds een grote voorkeur voor keuzes die tot een lagere kostprijs leiden. Hoe dan ook, de vraag naar duurzame producten neemt gestaag toe ondanks het feit dat deze producten toch vaak iets duurder zijn dan hun conventionele tegenhangers. Deze richting wordt ook meer en meer ondersteund door wetgeving en het prijskaartje van een duurzaam product wordt regelmatig kunstmatig verlaagd door allerlei soorten subsidies. Het is, ondanks deze stimulerende maatregelingen, echter nog steeds een hele uitdaging om de mate van duurzaamheid van bestaande producten te verbeteren. De huidige fabrieken zijn op een specifieke manier ontworpen om conventionele producten optimaal te produceren en dit beperkt vaak de mogelijkheden om ze (nog) duurzamer te produceren. Ook zijn klanten (vaak) gewend aan bepaalde soms secundaire productkenmerken. Bijvoorbeeld, in een aantal gevallen worden er nog specifieke additieven aan de latex toegevoegd voor een optimale prestatie en de alternatieve latex heeft mogelijk andere additieven nodig om vergelijkbare eigenschappen te krijgen als het oorspronkelijke product. Bovendien neemt de levensduur van producten snel af en is er een verschuiving gaande van grote, niet flexibele, productieprocessen naar kleinere flexibele tegenhangers. Kortom, voor het maken van commerciële interessante latexen zijn veel stappen nodig en elke stap kan een behoorlijk invloed hebben op het uiteindelijke nivo van duurzaamheid van het eind product. Het is dus ook van groot belang dat de hele productieketen wordt beoordeeld als het nivo van duurzaamheid van een nieuw product wordt bepaald.
Synthetische latexen zijn de basisingrediënten voor commercieel verkrijgbare lijmen, inkt en verven en er worden jaarlijks meer dan 10 miljoen ton wereldwijd geproduceerd. De latexen worden meestal bereid met behulp van een vrije radicaalpolymerisatie in semi-batch reactor. Deze manier van produceren vereist intensieve menging omdat de reagentia tijdens de verwerking worden toegevoegd en de polymerisatiereactie exotherm is. De controle van de reactietemperatuur en de mate van menging tijdens de verwerking is cruciaal en het controleniveau neemt toe met het oppervlak tot volumeverhouding van de gebruikte reactor. Een verkleining van de reactor valt daarom ook samen met een verbeterde controle van de reactie. Echter, het toepassen van een kleinere reactor is bij een (semi-) batch proces vanuit een economisch standpunt minder interessant. Overschakelen naar een kleine continue reactor is aantrekkelijk vanuit een commercieel oogpunt, maar vereist over het algemeen een aanzienlijke verandering in het ontwerp en de organisatie van de fabriek. Het feit dat de semi-batch werkwijze nog steeds gangbaar is in de emulsiepolymerisatie industrie, ondanks het al meer dan een decennium beschikbaar zijn van “continue” tegenhangers, suggereert dat deze omschakeling commercieel gezien nog niet aantrekkelijk genoeg is.
Zetmeel gestabiliseerde latexen zijn al sinds 1950 commercieel beschikbaar maar de tegenhangers op basis van synthetische stabilisatoren zijn op dit moment het meest gangbaar in de markten van verf en lijm. Zetmeel is een hernieuwbare en biologisch afbreekbare grondstof die veel beter in water oplost dan bijvoorbeeld cellulose. Dit maakt zetmeel niet alleen eenvoudiger biologisch afbreekbaar dan cellulose, maar ook veel meer geschikt voor modificatie en derivatisering in dit milieuvriendelijke oplosmiddel. Zetmeel
Samenvatting
128 129
kan afkomstig zijn van verschillende plantaardige bronnen waarvan maïs, tarwe, cassave (tapioca) en aardappel de belangrijkste zijn voor toepassing op industriële schaal. Elk type zetmeel heeft zijn eigen specifieke set van voor- en nadelen waardoor ze meer of minder interessant voor bepaalde toepassingen zijn. Aardappelzetmeel staat bekend om zijn lage vet- en eiwitgehalte en de amylopectine (meestal aangeduid met de term waxy) variant heeft zelfs een nog hogere homogeniteit vanwege het zeer lage gehalte aan amylose. De laatste is sinds kort op commerciële schaal beschikbaar en heeft ook een aantal unieke eigenschappen met betrekking tot viscositeit, associatief gedrag met hydrofobe groepen en stabiliteit na oplossen. Waxy aardappelzetmeel producten genereren een hogere viscositeit en stabiliteit dan de tegenhangers op basis van gewoon aardappelzetmeel en de laatste resulteert weer in aanzienlijk meer viscositeit en stabiliteit dan producten gemaakt van regulier graan of tapioca zetmeel.
Synthetische latexen worden meestal bereid in polymerisatiereactoren van enkele kubieke meters want dit is een commercieel breed geaccepteerde manier om goedkope producten te maken. De bijbehorende polymerisatie procedures zijn vaak al volledig geoptimaliseerd en het gebruik van goedkopere ingrediënten lijkt dan de meest eenvoudige manier om de kosten nog verder verlagen. Echter, een verandering in formulering heeft niet alleen invloed op de eigenschappen van het uiteindelijke product maar ook op de eigenschappen van het reactiemengsel. Het toepassen van goedkopere ingrediënten kan, in het ergste geval, tot inferieure producten en/of duurdere productieprocessen leiden. De impact van de aanpassing, op zowel het resterende deel van de formulering als bijbehorend polymerisatieproces, moet daarom ook eerst goed onderzocht worden. Verschillende proeven om de synthetische hulpstoffen in commerciële latexen te vervangen door gemodificeerd zetmeel zijn in het verleden al uitgevoerd maar het is nog niet gelukt om een commercieel aantrekkelijk alternatief te maken. Er zijn voldoende relevante publicaties beschikbaar (tot wel 60 jaar oud) die als uitgangspunt kunnen dienen om de optredende problemen op te lossen. Alleen is de beschikbare informatie enigszins achterhaald omdat de westerse samenleving drastisch is veranderd in de afgelopen decennia. Ten eerste, gezondheid en milieu gerelateerde onderwerpen zijn tegenwoordig bijvoorbeeld veel belangrijker geworden. Ten tweede, veel in het verleden gebruikte ingrediënten worden steeds schaarser of sommige additieven mogen nu niet meer worden gebruikt. Ten derde, hebben energie- en arbeidskosten nu een veel groter aandeel in de kostprijs dan vroeger. En tot slot, zijn gebruikers nu aan producten met een veel hoger kwaliteitsniveau gewend. Voor de beschikbare zetmeel gebaseerde latexen is dus ook extra onderzoek nodig voordat ze serieus kunnen concurreren met de huidige latexen die gestabiliseerd zijn met synthetische hulpstoffen. Ook is het belangrijk om alle stappen van de levenscyclus van het product te beoordelen en het onderzoek te richten op die toepassingen waarin de (chemische) modificatie van het zetmeel tot een minimum kunnen worden beperkt. Dit is niet alleen gunstig vanuit het oogpunt van kostprijs, maar ook voor het niveau van duurzaamheid en biologische afbreekbaarheid.
Er is een voorkeur voor toepassingen die een goede hechting tussen de latex deeltjes en cellulose gebaseerde materialen vereisen. Het toevoegen van zetmeel aan de betreffende
Productontwikkeling wordt aan de ene kant meer en meer beïnvloed door mens en milieu gerelateerde zaken maar aan de andere kant is er nog steeds een grote voorkeur voor keuzes die tot een lagere kostprijs leiden. Hoe dan ook, de vraag naar duurzame producten neemt gestaag toe ondanks het feit dat deze producten toch vaak iets duurder zijn dan hun conventionele tegenhangers. Deze richting wordt ook meer en meer ondersteund door wetgeving en het prijskaartje van een duurzaam product wordt regelmatig kunstmatig verlaagd door allerlei soorten subsidies. Het is, ondanks deze stimulerende maatregelingen, echter nog steeds een hele uitdaging om de mate van duurzaamheid van bestaande producten te verbeteren. De huidige fabrieken zijn op een specifieke manier ontworpen om conventionele producten optimaal te produceren en dit beperkt vaak de mogelijkheden om ze (nog) duurzamer te produceren. Ook zijn klanten (vaak) gewend aan bepaalde soms secundaire productkenmerken. Bijvoorbeeld, in een aantal gevallen worden er nog specifieke additieven aan de latex toegevoegd voor een optimale prestatie en de alternatieve latex heeft mogelijk andere additieven nodig om vergelijkbare eigenschappen te krijgen als het oorspronkelijke product. Bovendien neemt de levensduur van producten snel af en is er een verschuiving gaande van grote, niet flexibele, productieprocessen naar kleinere flexibele tegenhangers. Kortom, voor het maken van commerciële interessante latexen zijn veel stappen nodig en elke stap kan een behoorlijk invloed hebben op het uiteindelijke nivo van duurzaamheid van het eind product. Het is dus ook van groot belang dat de hele productieketen wordt beoordeeld als het nivo van duurzaamheid van een nieuw product wordt bepaald.
Synthetische latexen zijn de basisingrediënten voor commercieel verkrijgbare lijmen, inkt en verven en er worden jaarlijks meer dan 10 miljoen ton wereldwijd geproduceerd. De latexen worden meestal bereid met behulp van een vrije radicaalpolymerisatie in semi-batch reactor. Deze manier van produceren vereist intensieve menging omdat de reagentia tijdens de verwerking worden toegevoegd en de polymerisatiereactie exotherm is. De controle van de reactietemperatuur en de mate van menging tijdens de verwerking is cruciaal en het controleniveau neemt toe met het oppervlak tot volumeverhouding van de gebruikte reactor. Een verkleining van de reactor valt daarom ook samen met een verbeterde controle van de reactie. Echter, het toepassen van een kleinere reactor is bij een (semi-) batch proces vanuit een economisch standpunt minder interessant. Overschakelen naar een kleine continue reactor is aantrekkelijk vanuit een commercieel oogpunt, maar vereist over het algemeen een aanzienlijke verandering in het ontwerp en de organisatie van de fabriek. Het feit dat de semi-batch werkwijze nog steeds gangbaar is in de emulsiepolymerisatie industrie, ondanks het al meer dan een decennium beschikbaar zijn van “continue” tegenhangers, suggereert dat deze omschakeling commercieel gezien nog niet aantrekkelijk genoeg is.
Zetmeel gestabiliseerde latexen zijn al sinds 1950 commercieel beschikbaar maar de tegenhangers op basis van synthetische stabilisatoren zijn op dit moment het meest gangbaar in de markten van verf en lijm. Zetmeel is een hernieuwbare en biologisch afbreekbare grondstof die veel beter in water oplost dan bijvoorbeeld cellulose. Dit maakt zetmeel niet alleen eenvoudiger biologisch afbreekbaar dan cellulose, maar ook veel meer geschikt voor modificatie en derivatisering in dit milieuvriendelijke oplosmiddel. Zetmeel
Samenvatting
130 131
pyrodextrinering als de corresponderende polymerisatie te kunnen bepalen.
De mate van duurzaamheid van een latex kan worden verbeterd door het droge stof gehalte te verhogen, energieverbruik tijdens de verwerking te verminderen of combinaties hiervan. Hoofdstuk 4 toont de invloed van veranderingen op dit gebied voor een polyvinylacetaat latex gestabiliseerd met enzymatische gemodificeerd waxy aardappelzetmeel (25 gew % op monomeer). Reactietemperaturen in de orde van 75 tot 85 °C waren geselecteerd en de concentratie van de initiator werd gevarieerd van 2 tot 4 gew %. Een gerandomiseerd 22 factorial design, aangevuld met vijf middelpunten, was geselecteerd om de invloed van deze twee factoren op de verschillende product en proces gerelateerde variabelen te onderzoeken. Aanzienlijke verschillen in reactieomstandigheden werden aangelegd maar alle polymerisaties waren waarschijnlijk uitgevoerd rond het kookpunt van de azeotroop vinylacetaat monomeer en water (66 °C). De viscositeit van de verkregen latexen varieerde van 700 tot 2300 mPa•s en een statistisch model werd verkregen in een poging om de latex viscositeit te correleren met de variabelen reactietemperatuur en de concentratie van de initiator.
Emulgatoren genereren hydrofobe aggregaten na oplossen in water en het vinylacetaat monomeer kan worden opgeslagen in deze aggregaten als deze in overmaat aanwezig is. De aanwezigheid van deze aggregaten verbetert de mate van beheersing van de polymerisatie met betrekking tot niveau van refluxen en deeltjesvorming. Octenyl gesuccinyleerd zetmeel heeft emulgerende eigenschappen en de invloed van dit type derivaat op het polymerisatieproces en latex eigenschappen wordt onderzocht in hoofdstuk 5. De zetmeelproducten waren geëxtrudeerd (een proces dat het zetmeel mechanisch bewerkt) voor gebruik, met het oog op het invoeren van een bescheiden niveau van afbraak. Deze behandeling maakt de producten ook koud water oplosbaar. Het formaat van de zetmeelfragmenten zijn na oplossen groot in verhouding tot de gangbare gebruikte en aanbevolen zetmeelderivaten voor de toepassing in emulsie polymerisatie (zetmeel wordt in deze toepassing meestal afgebroken tot het nivo van oligomeren. Latexen gebaseerd op waxy aardappelzetmeel vertoonden de meest opmerkelijke verandering na een octenyl succinylatie. Een toename van ~ 1000 tot ~ 10000 mPa•s werd waargenomen voor de waxy variant met een substitutiegraad van 0.02 terwijl de reguliere tegenhanger slechts tot een stijging van ~ 1000 tot ~ 3000 mPa•s leidt.
Het aanbrengen van lijm met rollers is een zeer populaire techniek in de houtbewerking industrie en vereist een viscositeit van minimaal duizend mPa•s. Een aantal van de gemaakte (octenyl gesuccinyleerde) (waxy) aardappelzetmeel gestabiliseerd latexen voldoen aan dit criterium en drie daarvan werden gecontroleerd op hout verlijmingeigenschappen in hoofdstuk 6. Het onderzoek heeft standaard houtlijm test “ EN204 D2 “ als uitgangspunt en deze test is representatief voor houtlijmtoepassingen in gebieden met slechts beperkte blootstelling aan vocht (bijvoorbeeld: keuken). De drie beoordeelde latexen hadden een houtbinding in de orde van 17 tot 24 MPa en dat is veel hoger dan de 8 MPa criterium van de geselecteerde test om de houtbinding te bepalen.
latex kan dan namelijk leiden tot synergetische effecten in de hechting want zetmeel heeft een hoge affiniteit met cellulose. Met deze gedachte zijn latex gebaseerde verven, inkten, coatingen en lijmen voor hout en papier gebaseerde materialen het meest interessant om te onderzoeken. Het beste uitgangspunt voor dit onderzoek lijken houtlijmen te zijn met polyvinylacetaat als belangrijkste ingrediënt. Enerzijds, zetmeel is een potentieel alternatief voor de gebruikte schutcolloïd (te weten: polyvinylalcohol) die vaak wordt gebruikt in dit type lijm. Anderzijds, de compatibiliteit tussen zetmeel en polyvinylacetaat is goed. En tot slot, de houtlijm markt is, in potentie, groot genoeg om een grondig onderzoek in de vervanging van het polyvinylalcohol gedeelte door zetmeel te rechtvaardigen.
Verschillende procedures voor het maken van polyvinylacetaat latex werden gevonden na een evaluatie van algemeen toegankelijke informatie. De verzamelde bereidingsprocedures werden vervolgens niet alleen uitgeprobeerd in de praktijk, maar ook geëvalueerd volgens de principes van de groene chemie - een leidraad om de mate van duurzaamheid van synthetische producten te verhogen. De verkregen polymerisatiekennis werd gebruikt om het systeem en de procedure dat in hoofdstuk 2 beschreven staat te ontwerpen. In dit hoofdstuk werd enzymatisch omgezet waxy aardappelzetmeel (10 gew. % op monomeer ) gebruikt als stabilisator in afwezigheid van synthetische toevoegingen (bijvoorbeeld emulgatoren, anti - schuim en schutcolloïd). De bijbehorende polymerisaties werden uitgevoerd zonder de verwijdering van remmers (zoals zuurstof en hydroquinone) uit de gebruikte grondstoffen en de invloed van drie onafhankelijke variabelen (te weten: niveau van menging, voordosering van monomeer en voordosering van een mengsel initiator en buffer) werden beoordeeld. De variabelen warmtetoevoer tijdens verwerking, monomeeromzetting, latex opbrengst, anion concentratie, pH, viscositeit, deeltjesgrootteverdeling, hoeveelheid geënt schutcolloïd en glasovergangstemperatuur werden bepaald. De reproduceerbaarheid van de opgestelde bereidingsprocedure was goed met betrekking tot de variatie waargenomen bij aanzienlijk verschillende instellingen. De verkregen resultaten bevatten tal van aanwijzingen die nuttig kunnen zijn als de deze latexen moeten worden geoptimaliseerd met betrekking tot de functionaliteit en het niveau van duurzaamheid.
De pyrodextrinering van zetmeel wordt voornamelijk beinvloed door de hoeveelheid toegevoegde warmte, vochtniveau en zoutzuur concentratie. Hoofdstuk 3 illustreert de invloed van veranderingen in de reactiecondities op de eigenschappen van aardappelzetmeel gebaseerde pyrodextrines en hun stabiliserende eigenschappen tijdens een emulsiepolymerisatie. Veranderingen in de zoutzuurconcentratie en de toegevoegde hoeveelheid warmte hadden een grotere impact op de stabiliserende eigenschappen van de pyrodextrines dan de verschillen in niveau van voordrogen. De pyrodextrines met het hoogste energieverbruik tijdens het pyrodextrineringsproces resulteerden in latexen met de meest gewenste latex kenmerken (te weten: monodisperse deeltjesgrootteverdeling). Daarnaast was er ook meer energie nodig om deze latexen te maken doordat er een behoorlijke reflux (dwz verspilling van energie) optrad tijdens de verwerking. Een aanzienlijke vermindering van het energieverbruik tijdens het maken van dit soort latexen is zeer waarschijnlijk haalbaar, maar aanvullend onderzoek is nodig om de optimale procescondities voor zowel de
Samenvatting
130 131
pyrodextrinering als de corresponderende polymerisatie te kunnen bepalen.
De mate van duurzaamheid van een latex kan worden verbeterd door het droge stof gehalte te verhogen, energieverbruik tijdens de verwerking te verminderen of combinaties hiervan. Hoofdstuk 4 toont de invloed van veranderingen op dit gebied voor een polyvinylacetaat latex gestabiliseerd met enzymatische gemodificeerd waxy aardappelzetmeel (25 gew % op monomeer). Reactietemperaturen in de orde van 75 tot 85 °C waren geselecteerd en de concentratie van de initiator werd gevarieerd van 2 tot 4 gew %. Een gerandomiseerd 22 factorial design, aangevuld met vijf middelpunten, was geselecteerd om de invloed van deze twee factoren op de verschillende product en proces gerelateerde variabelen te onderzoeken. Aanzienlijke verschillen in reactieomstandigheden werden aangelegd maar alle polymerisaties waren waarschijnlijk uitgevoerd rond het kookpunt van de azeotroop vinylacetaat monomeer en water (66 °C). De viscositeit van de verkregen latexen varieerde van 700 tot 2300 mPa•s en een statistisch model werd verkregen in een poging om de latex viscositeit te correleren met de variabelen reactietemperatuur en de concentratie van de initiator.
Emulgatoren genereren hydrofobe aggregaten na oplossen in water en het vinylacetaat monomeer kan worden opgeslagen in deze aggregaten als deze in overmaat aanwezig is. De aanwezigheid van deze aggregaten verbetert de mate van beheersing van de polymerisatie met betrekking tot niveau van refluxen en deeltjesvorming. Octenyl gesuccinyleerd zetmeel heeft emulgerende eigenschappen en de invloed van dit type derivaat op het polymerisatieproces en latex eigenschappen wordt onderzocht in hoofdstuk 5. De zetmeelproducten waren geëxtrudeerd (een proces dat het zetmeel mechanisch bewerkt) voor gebruik, met het oog op het invoeren van een bescheiden niveau van afbraak. Deze behandeling maakt de producten ook koud water oplosbaar. Het formaat van de zetmeelfragmenten zijn na oplossen groot in verhouding tot de gangbare gebruikte en aanbevolen zetmeelderivaten voor de toepassing in emulsie polymerisatie (zetmeel wordt in deze toepassing meestal afgebroken tot het nivo van oligomeren. Latexen gebaseerd op waxy aardappelzetmeel vertoonden de meest opmerkelijke verandering na een octenyl succinylatie. Een toename van ~ 1000 tot ~ 10000 mPa•s werd waargenomen voor de waxy variant met een substitutiegraad van 0.02 terwijl de reguliere tegenhanger slechts tot een stijging van ~ 1000 tot ~ 3000 mPa•s leidt.
Het aanbrengen van lijm met rollers is een zeer populaire techniek in de houtbewerking industrie en vereist een viscositeit van minimaal duizend mPa•s. Een aantal van de gemaakte (octenyl gesuccinyleerde) (waxy) aardappelzetmeel gestabiliseerd latexen voldoen aan dit criterium en drie daarvan werden gecontroleerd op hout verlijmingeigenschappen in hoofdstuk 6. Het onderzoek heeft standaard houtlijm test “ EN204 D2 “ als uitgangspunt en deze test is representatief voor houtlijmtoepassingen in gebieden met slechts beperkte blootstelling aan vocht (bijvoorbeeld: keuken). De drie beoordeelde latexen hadden een houtbinding in de orde van 17 tot 24 MPa en dat is veel hoger dan de 8 MPa criterium van de geselecteerde test om de houtbinding te bepalen.
latex kan dan namelijk leiden tot synergetische effecten in de hechting want zetmeel heeft een hoge affiniteit met cellulose. Met deze gedachte zijn latex gebaseerde verven, inkten, coatingen en lijmen voor hout en papier gebaseerde materialen het meest interessant om te onderzoeken. Het beste uitgangspunt voor dit onderzoek lijken houtlijmen te zijn met polyvinylacetaat als belangrijkste ingrediënt. Enerzijds, zetmeel is een potentieel alternatief voor de gebruikte schutcolloïd (te weten: polyvinylalcohol) die vaak wordt gebruikt in dit type lijm. Anderzijds, de compatibiliteit tussen zetmeel en polyvinylacetaat is goed. En tot slot, de houtlijm markt is, in potentie, groot genoeg om een grondig onderzoek in de vervanging van het polyvinylalcohol gedeelte door zetmeel te rechtvaardigen.
Verschillende procedures voor het maken van polyvinylacetaat latex werden gevonden na een evaluatie van algemeen toegankelijke informatie. De verzamelde bereidingsprocedures werden vervolgens niet alleen uitgeprobeerd in de praktijk, maar ook geëvalueerd volgens de principes van de groene chemie - een leidraad om de mate van duurzaamheid van synthetische producten te verhogen. De verkregen polymerisatiekennis werd gebruikt om het systeem en de procedure dat in hoofdstuk 2 beschreven staat te ontwerpen. In dit hoofdstuk werd enzymatisch omgezet waxy aardappelzetmeel (10 gew. % op monomeer ) gebruikt als stabilisator in afwezigheid van synthetische toevoegingen (bijvoorbeeld emulgatoren, anti - schuim en schutcolloïd). De bijbehorende polymerisaties werden uitgevoerd zonder de verwijdering van remmers (zoals zuurstof en hydroquinone) uit de gebruikte grondstoffen en de invloed van drie onafhankelijke variabelen (te weten: niveau van menging, voordosering van monomeer en voordosering van een mengsel initiator en buffer) werden beoordeeld. De variabelen warmtetoevoer tijdens verwerking, monomeeromzetting, latex opbrengst, anion concentratie, pH, viscositeit, deeltjesgrootteverdeling, hoeveelheid geënt schutcolloïd en glasovergangstemperatuur werden bepaald. De reproduceerbaarheid van de opgestelde bereidingsprocedure was goed met betrekking tot de variatie waargenomen bij aanzienlijk verschillende instellingen. De verkregen resultaten bevatten tal van aanwijzingen die nuttig kunnen zijn als de deze latexen moeten worden geoptimaliseerd met betrekking tot de functionaliteit en het niveau van duurzaamheid.
De pyrodextrinering van zetmeel wordt voornamelijk beinvloed door de hoeveelheid toegevoegde warmte, vochtniveau en zoutzuur concentratie. Hoofdstuk 3 illustreert de invloed van veranderingen in de reactiecondities op de eigenschappen van aardappelzetmeel gebaseerde pyrodextrines en hun stabiliserende eigenschappen tijdens een emulsiepolymerisatie. Veranderingen in de zoutzuurconcentratie en de toegevoegde hoeveelheid warmte hadden een grotere impact op de stabiliserende eigenschappen van de pyrodextrines dan de verschillen in niveau van voordrogen. De pyrodextrines met het hoogste energieverbruik tijdens het pyrodextrineringsproces resulteerden in latexen met de meest gewenste latex kenmerken (te weten: monodisperse deeltjesgrootteverdeling). Daarnaast was er ook meer energie nodig om deze latexen te maken doordat er een behoorlijke reflux (dwz verspilling van energie) optrad tijdens de verwerking. Een aanzienlijke vermindering van het energieverbruik tijdens het maken van dit soort latexen is zeer waarschijnlijk haalbaar, maar aanvullend onderzoek is nodig om de optimale procescondities voor zowel de
132
APPENDICESSummary SamenvattingDankwoordPublications
132
APPENDICESSummary SamenvattingDankwoordPublications
Dankwoord
134 135
proefschrift vond ik wel heel erg leerzaam. Het nut van de academische stijl van rapporteren en presenteren staat wat mij betreft dus ook niet ter discussie en ik zal er in de toekomst absoluut vaker gebruik van gaan maken.
In 2011 werd duidelijk dat het onderwerp waar ik mee bezig was in aanmerking kwam voor de Kenniswerkersregeling en daarmee was de financiering rond voor een aanzienlijke uitbreiding. Sinds die tijd is Guus Maas de penvoerder in het project “Hernieuwbare polymeeremulsies” waar ook mijn promotieonderzoek een onderdeel van werd. Ik heb intussen samen met Guus al heel wat onderwerpen uitvoerig bediscussieerd. Guus, bedankt voor je ondersteuning en het feit dat je altijd tijd voor mij hebt vrij gemaakt als ik weer eens iets van mij moest afpraten. Aly Boer en Anna Marie Dijk wil ik bedanken voor het delen van jullie kennis en vaardigheden met betrekking op synthetische lijmen in de toepassing. De door jullie aangeleverde informatie is in feite het fundament van hoofdstuk 6 in dit proefschrift. Ook wil ik graag mijn waardering uitspreken voor de collegialiteit en inzet van mijn laboratoriumgenoten Egbert Hadderingh, Marcel Staal, Henk Westra, Renaldo ter Stege, Sander Schuiling, Jeanette Dekker en Erin Paske. Niet alle zetmeelderivaten die jullie voor mij gemaakt hebben zijn terug te vinden in mijn proefschrift maar de betreffende producten hebben zeker een rol van betekenis in het onderzoek dat buiten dit proefschrift om nog steeds gaande is. Paul Nommensen heeft een heel mooi werkend Java-programma voor mij geschreven waarmee het secundaire deeltjesvormingsproces in de radicaalpolymerisatie kan worden gemodelleerd. Paul, ik ben erg tevreden over wat je voor mij hebt gemaakt maar ik had helaas niet genoeg tijd om dit onderdeel in mijn proefschrift op te nemen. Zonder goede analyses heb je niet voldoende informatie om het onderzoek in de juiste richting te laten gaan en in dit kader wil ik mijn collega’s Jan Hulshof (deeltjesgrooteverdelingen), Bert Smit en Wim Akkerman (Anionen) en Menko van der Ark, Wubbo Mulder en Fatma Asdarci (ethanal en vinylacetaat) bedanken. Zetmeel gerelateerde patenten zijn zeer complex doordat het gebruik van zetmeel al zeer intensief onderzocht en gepatenteerd is. In dit licht wil ik Piet Buwalda, Henk Meima en Carla Kuiper bedanken voor het delen van jullie kennis en visie met betrekking tot de wereld van de patenten. Van de afdeling marketing wil ik Robert Krol, Bruno Vanpoucke, Joep Staats en Vincent Lamberti nog even noemen. We hebben tijdens de uitvoering van het onderzoek niet vaak overleg gehad maar we zijn nu wel op zoek naar mogelijkheden om de vergaarde kennis onder de aandacht te brengen bij (potentiele) klanten. Hopelijk kunnen we straks de commercieële vruchten van dit onderzoek gaan plukken.
Begin 2012 werd onze subsidieaanvraag in het kader van “Verordening Transitie II en Pieken“ toegekend door het samenwerkingsverband Noord-Nederland en de provincie Groningen. De bijbehorende financiering maakte een verdere uitbreiding van het project “Hernieuwbare polymeer emulsies” mogelijk en dit leidde tot de aanstelling van Louis Daniel als postdoc voor de duur van 2 jaar. Louis, your assistance with the execution of the wood glue tests made it possible to finish chapter 6 just in time. From my point of view, this dissertation would not have been complete without this chapter. I appreciate your contribution (i.e. grafting related experiments and wood glue tests) and good luck with your next post-doc assignment at the university of Groningen. Ik wil ook mijn dank uitspreken aan de medewerkers, studenten,
Een generalist die de theorie bij de praktijk zoekt. Zo zie ik mijzelf en dat al ruim 20 jaar. Het was voor mij dus ook logisch dat ik na mijn HBO-opleiding ben gaan werken in plaats van door te stromen naar de universiteit. Ik heb in die periode nog wel een aantal keren overwogen om een academische graad te gaan halen. Dit om de kans op een leuke baan te vergroten want halverwege de jaren 90 waren er in mijn vakgebied meer vacatures voor academici dan voor HBO’ers. Ik heb echter nooit lang zonder baan gezeten en daardoor is de noodzaak vanuit deze hoek er nooit geweest. Bovendien had ik in die tijd ook helemaal geen behoefte aan een tijdrovende studie want mijn vrije tijd ging voornamelijk op aan het trainen van paarden voor springwedstrijden. Tot het jaar 2007, in dit jaar werd mij duidelijk dat ik mijn academische vaardigheden moest verbeteren om mijn werk beter uit te kunnen voeren. Het heeft uiteindelijk, door werk en privé gerelateerde omstandigheden, nog wel tot 2010 geduurd voordat ik echt aan de slag kon met het onderzoek dat tot dit proefschrift heeft geleid.
Mijn dank gaat in de eerste plaats uit naar Johan Hopman. Ik weet nog precies waar ik stond (tot aan mijn knieën in het water van een riviertje in het Sauerland), wat voor weer het was (lekker nazomer weer), wat ik deed (vliegvissen), met wie (mijn broer) en wat ik dacht (jaja….) toen je me in 2005 belde met de mededeling dat de AVEBE mij (opnieuw) overbodig had verklaard, maar dat je er alles aan zou doen om mij binnen de organisatie te houden. Bijna 10 jaar later is duidelijk geworden dat je niet alleen je best hebt gedaan om mij binnen de organisatie te houden maar ook om mijn persoonlijke ontwikkeling te ondersteunen met raad en daad. Anne Margriet Hofman wil ik als volgende bedanken. Tijdens mijn promotie ben ik meerdere keren tot de conclusie gekomen dat mijn promotieonderzoek niet was te combineren met mijn reguliere AVEBE werk en mijn moeilijke thuissituatie. Als ik met deze boodschap bij jou langs kwam was de strekking van jouw reactie (gelukkig) altijd “Ok, doe nu dan eerst maar wat wel mogelijk is en dan regel ik de rest” en is ook een belangrijke reden waarom ik nu een dankwoord voor mijn proefschrift aan het schrijven ben.
Promoveren is natuurlijk alleen mogelijk als je promotoren kunt vinden die het vertrouwen in je uitspreken en bereid zijn om je te willen helpen om je doel te bereiken. Dit gaat ook op als je een HBO’er bent, voornamelijk ervaring hebt opgedaan in commerciële laboratoria en ook nog eens wil gaan promoveren in een onderwerp die door de academische wereld al enigszins verlaten is. Professor Francesco Picchioni, bedankt dat je mij tijdens het onderzoek hebt gecoacht en dat je mijn eerste promotor hebt willen zijn. Jouw inzet en enthousiasme hebben mij geholpen om de periodes door te komen waarin ik het gevoel had dat het mij allemaal wel heel erg tegen zat. Ook waardeer ik het feit dat je altijd tijd vrij voor mij hebt gemaakt als dat nodig was. Niet alleen om mijn manier van werken beter af te stemmen op die van de academische wereld maar ook voor gesprekken op een persoonlijk vlak. Sinds het afgelopen jaar is het woord leesbaarheid voor mij sterk verbonden met mijn tweede promotor Professor Erik Heeres. Erik, het was wel even slikken toen je 2 weken voor het einde van mijn promotietraject aangaf dat ik alle hoofdstukken opnieuw moest gaan indelen. Ook heeft het even geduurd voordat ik begreep wat je nu werkelijk verstond onder de term leesbaarheid. Persoonlijk ben ik nog niet helemaal overtuigd van de noodzaak van de doorgevoerde aanpassingen maar het samen met jou opnieuw indelen van verschillende onderdelen van het
Dankwoord
134 135
proefschrift vond ik wel heel erg leerzaam. Het nut van de academische stijl van rapporteren en presenteren staat wat mij betreft dus ook niet ter discussie en ik zal er in de toekomst absoluut vaker gebruik van gaan maken.
In 2011 werd duidelijk dat het onderwerp waar ik mee bezig was in aanmerking kwam voor de Kenniswerkersregeling en daarmee was de financiering rond voor een aanzienlijke uitbreiding. Sinds die tijd is Guus Maas de penvoerder in het project “Hernieuwbare polymeeremulsies” waar ook mijn promotieonderzoek een onderdeel van werd. Ik heb intussen samen met Guus al heel wat onderwerpen uitvoerig bediscussieerd. Guus, bedankt voor je ondersteuning en het feit dat je altijd tijd voor mij hebt vrij gemaakt als ik weer eens iets van mij moest afpraten. Aly Boer en Anna Marie Dijk wil ik bedanken voor het delen van jullie kennis en vaardigheden met betrekking op synthetische lijmen in de toepassing. De door jullie aangeleverde informatie is in feite het fundament van hoofdstuk 6 in dit proefschrift. Ook wil ik graag mijn waardering uitspreken voor de collegialiteit en inzet van mijn laboratoriumgenoten Egbert Hadderingh, Marcel Staal, Henk Westra, Renaldo ter Stege, Sander Schuiling, Jeanette Dekker en Erin Paske. Niet alle zetmeelderivaten die jullie voor mij gemaakt hebben zijn terug te vinden in mijn proefschrift maar de betreffende producten hebben zeker een rol van betekenis in het onderzoek dat buiten dit proefschrift om nog steeds gaande is. Paul Nommensen heeft een heel mooi werkend Java-programma voor mij geschreven waarmee het secundaire deeltjesvormingsproces in de radicaalpolymerisatie kan worden gemodelleerd. Paul, ik ben erg tevreden over wat je voor mij hebt gemaakt maar ik had helaas niet genoeg tijd om dit onderdeel in mijn proefschrift op te nemen. Zonder goede analyses heb je niet voldoende informatie om het onderzoek in de juiste richting te laten gaan en in dit kader wil ik mijn collega’s Jan Hulshof (deeltjesgrooteverdelingen), Bert Smit en Wim Akkerman (Anionen) en Menko van der Ark, Wubbo Mulder en Fatma Asdarci (ethanal en vinylacetaat) bedanken. Zetmeel gerelateerde patenten zijn zeer complex doordat het gebruik van zetmeel al zeer intensief onderzocht en gepatenteerd is. In dit licht wil ik Piet Buwalda, Henk Meima en Carla Kuiper bedanken voor het delen van jullie kennis en visie met betrekking tot de wereld van de patenten. Van de afdeling marketing wil ik Robert Krol, Bruno Vanpoucke, Joep Staats en Vincent Lamberti nog even noemen. We hebben tijdens de uitvoering van het onderzoek niet vaak overleg gehad maar we zijn nu wel op zoek naar mogelijkheden om de vergaarde kennis onder de aandacht te brengen bij (potentiele) klanten. Hopelijk kunnen we straks de commercieële vruchten van dit onderzoek gaan plukken.
Begin 2012 werd onze subsidieaanvraag in het kader van “Verordening Transitie II en Pieken“ toegekend door het samenwerkingsverband Noord-Nederland en de provincie Groningen. De bijbehorende financiering maakte een verdere uitbreiding van het project “Hernieuwbare polymeer emulsies” mogelijk en dit leidde tot de aanstelling van Louis Daniel als postdoc voor de duur van 2 jaar. Louis, your assistance with the execution of the wood glue tests made it possible to finish chapter 6 just in time. From my point of view, this dissertation would not have been complete without this chapter. I appreciate your contribution (i.e. grafting related experiments and wood glue tests) and good luck with your next post-doc assignment at the university of Groningen. Ik wil ook mijn dank uitspreken aan de medewerkers, studenten,
Een generalist die de theorie bij de praktijk zoekt. Zo zie ik mijzelf en dat al ruim 20 jaar. Het was voor mij dus ook logisch dat ik na mijn HBO-opleiding ben gaan werken in plaats van door te stromen naar de universiteit. Ik heb in die periode nog wel een aantal keren overwogen om een academische graad te gaan halen. Dit om de kans op een leuke baan te vergroten want halverwege de jaren 90 waren er in mijn vakgebied meer vacatures voor academici dan voor HBO’ers. Ik heb echter nooit lang zonder baan gezeten en daardoor is de noodzaak vanuit deze hoek er nooit geweest. Bovendien had ik in die tijd ook helemaal geen behoefte aan een tijdrovende studie want mijn vrije tijd ging voornamelijk op aan het trainen van paarden voor springwedstrijden. Tot het jaar 2007, in dit jaar werd mij duidelijk dat ik mijn academische vaardigheden moest verbeteren om mijn werk beter uit te kunnen voeren. Het heeft uiteindelijk, door werk en privé gerelateerde omstandigheden, nog wel tot 2010 geduurd voordat ik echt aan de slag kon met het onderzoek dat tot dit proefschrift heeft geleid.
Mijn dank gaat in de eerste plaats uit naar Johan Hopman. Ik weet nog precies waar ik stond (tot aan mijn knieën in het water van een riviertje in het Sauerland), wat voor weer het was (lekker nazomer weer), wat ik deed (vliegvissen), met wie (mijn broer) en wat ik dacht (jaja….) toen je me in 2005 belde met de mededeling dat de AVEBE mij (opnieuw) overbodig had verklaard, maar dat je er alles aan zou doen om mij binnen de organisatie te houden. Bijna 10 jaar later is duidelijk geworden dat je niet alleen je best hebt gedaan om mij binnen de organisatie te houden maar ook om mijn persoonlijke ontwikkeling te ondersteunen met raad en daad. Anne Margriet Hofman wil ik als volgende bedanken. Tijdens mijn promotie ben ik meerdere keren tot de conclusie gekomen dat mijn promotieonderzoek niet was te combineren met mijn reguliere AVEBE werk en mijn moeilijke thuissituatie. Als ik met deze boodschap bij jou langs kwam was de strekking van jouw reactie (gelukkig) altijd “Ok, doe nu dan eerst maar wat wel mogelijk is en dan regel ik de rest” en is ook een belangrijke reden waarom ik nu een dankwoord voor mijn proefschrift aan het schrijven ben.
Promoveren is natuurlijk alleen mogelijk als je promotoren kunt vinden die het vertrouwen in je uitspreken en bereid zijn om je te willen helpen om je doel te bereiken. Dit gaat ook op als je een HBO’er bent, voornamelijk ervaring hebt opgedaan in commerciële laboratoria en ook nog eens wil gaan promoveren in een onderwerp die door de academische wereld al enigszins verlaten is. Professor Francesco Picchioni, bedankt dat je mij tijdens het onderzoek hebt gecoacht en dat je mijn eerste promotor hebt willen zijn. Jouw inzet en enthousiasme hebben mij geholpen om de periodes door te komen waarin ik het gevoel had dat het mij allemaal wel heel erg tegen zat. Ook waardeer ik het feit dat je altijd tijd vrij voor mij hebt gemaakt als dat nodig was. Niet alleen om mijn manier van werken beter af te stemmen op die van de academische wereld maar ook voor gesprekken op een persoonlijk vlak. Sinds het afgelopen jaar is het woord leesbaarheid voor mij sterk verbonden met mijn tweede promotor Professor Erik Heeres. Erik, het was wel even slikken toen je 2 weken voor het einde van mijn promotietraject aangaf dat ik alle hoofdstukken opnieuw moest gaan indelen. Ook heeft het even geduurd voordat ik begreep wat je nu werkelijk verstond onder de term leesbaarheid. Persoonlijk ben ik nog niet helemaal overtuigd van de noodzaak van de doorgevoerde aanpassingen maar het samen met jou opnieuw indelen van verschillende onderdelen van het
Dankwoord
136 137
tijd, energie en creativiteit hebt willen steken in het maken van de omslag van dit proefschrift. Mooi dat het je gelukt is om mijn twee passies (paarden(sport) en laboratoriumonderzoek) met elkaar hebt kunnen vereenigen. Arjan, ook nog bedankt voor het accepteren van mijn verzoek om één van mijn paranimfen te zijn. Tessa, we kennen elkaar nog niet zo heel erg lang maar wel zo goed dat ik je gevraagd heb om één van mijn paranimfen te zijn. Ik stel het niet alleen erg op prijs dat je dit verzoek hebt aanvaard maar ook dat je mij in de afgelopen maanden op veel vlakken hebt geholpen. Jouw adviezen waarmee mijn verzameling zinnen met informatie meer een verhaal (kunnen) worden zullen me zeker in de toekomst helpen om de boodschap van mijn onderzoek beter over te brengen.
Ik heb erg veel plezier in mijn werk maar ik zou het niet kunnen volhouden zonder een intensieve hobby voor de afwisseling. Mijn dank gaat daarom ook uit naar de ruiters en amazones van de Kajützhoeve. Mede dankzij jullie inzet en enthousiasme bestaat onze manege nog steeds en kan ik nog steeds intensief bezig zijn met de paardensport op een manier die ik heel erg leuk vind. Jarna, Onno en Paul, we spreken elkaar soms tijden niet maar als ik jullie tegen kom dan heb ik het altijd gevoel dat we elkaar de dag ervoor nog hebben gesproken. Ook in de afgelopen moeilijke jaren hebben jullie weer bewezen dat jullie er op een of andere manier altijd voor mij zijn als het er echt op aankomt. Heel erg bedankt hiervoor.
promovendi en postdocs van de sectie scheikundige technologie van de universiteit van Groningen die mij in de afgelopen periode gastvrij hebben ontvangen en op weg hebben geholpen. In het bijzonder de medewerkers van de technische dienst (Marcel de Vries en Anne Appeldoorn) en de glasblazerij (Maarten Vervoort) want zonder jullie expertise en hulp had ik de configuratie van polymerisatiereactor niet kunnen optimaliseren gedurende het onderzoek. Arjan Kloekhorst wil ik bedanken voor het maken van de schematische weergave van de polymerisatie reactor en die terug te vinden is in meerdere hoofdstukken. Daarnaast heb ik met veel plezier Sunny Qian begeleidt met haar afstudeeropdracht waarmee ze haar master op de universiteit (cum laude) kon afronden. Gert Alberda van Ekenstein heeft niet alleen mijn kennis met betrekking op de DSC op een nivo hoger gezet maar ook nog alle DSC-metingen voor mij uitgevoerd op de universiteit. I also want to thank the reading committee of my dissertation. Prof. dr. A.A. Broekhuis, prof. dr. K.U. Loos and prof. dr. L. Moscicki, thank you for taking the time and effort to read my thesis. Your recommendations with respect to the technological assessment were used to improve this section considerably. Met de in 2012 toegekende subsidie kwam er ook geld beschikbaar om een Labview programma te laten schrijven waarmee de apparatuur van de laboratoriumopstelling kon worden aangestuurd. Dank zij Carya Automatisering (André Buurman en Paul Rijkers) werd de aansturing en monitoring van de polymerisatieopstelling, na een intensieve samenwerking, een stuk eenvoudiger en had ik vanaf dat moment (hoofdstuk 4) maar 1 softwareprogramma nodig (in plaats van 6).
Een deel van het proefschrift is te herleiden naar kennis en vaardigheden die ik mij voor 2010 eigen heb gemaakt. Ik wil daarom mijn (oud-)collega’s van het Nederlands Instituut voor Koolhydraat Onderzoek (NIKO), Packard Bioscience, Perkin Elmer, DSM Biologics, TNO Quality of life en AVEBE dus ook bedanken voor het creëren van een (over het algemeen) zeer prettige en leerzame werkomgeving. In dit kader wil ik in het bijzonder Harry van Lune nog even noemen. Harry, ik weet niet of het nog voor jou zo 123 herkenbaar is maar mijn manier van werken is nog steeds te herleiden naar ervaring die wij samen hebben opgedaan bij Packard Bioscience.
Eigenlijk zou mijn promotietraject in 2008 beginnen maar dat werd uitgesteld omdat ik mee wou helpen om een groot probleem in de semi-droge crosslink lijn van de extruderfabriek op te lossen. In augustus 2009 nam ik mij stellig voor dat vanaf dat moment alles moest wijken voor een goede start van het promotieonderzoek. Het liep echter totaal anders. Begin september bleek Marjan ernstig ziek te zijn en kwam alles voornamelijk in het teken te staan van zeer ingrijpende en langdurige medische behandelingen en het leren omgaan met de nieuwe situatie. Marjan heeft uiteindelijk altijd begrip op weten te brengen voor mijn situatie en dit is een zeer belangrijke reden waarom ik in deze moeilijke periode niet ben gestopt met het promotieonderzoek. Ik ben heel erg blij dat we nog steeds hele goede vrienden zijn ondanks het feit dat we nu niet meer samenwonen. Pap, Mam, Arjan, we hebben intussen samen al heel wat moeilijkheden doorstaan en daarin hebben wij elkaar altijd geholpen waar het maar even mogelijk was. Voor ons is dat vanzelfsprekend maar intussen is het mij wel duidelijk dat deze onderlinge verhouding lang niet in elk gezin aanwezig is. Pap, ook nog bedankt dat je
Dankwoord
136 137
tijd, energie en creativiteit hebt willen steken in het maken van de omslag van dit proefschrift. Mooi dat het je gelukt is om mijn twee passies (paarden(sport) en laboratoriumonderzoek) met elkaar hebt kunnen vereenigen. Arjan, ook nog bedankt voor het accepteren van mijn verzoek om één van mijn paranimfen te zijn. Tessa, we kennen elkaar nog niet zo heel erg lang maar wel zo goed dat ik je gevraagd heb om één van mijn paranimfen te zijn. Ik stel het niet alleen erg op prijs dat je dit verzoek hebt aanvaard maar ook dat je mij in de afgelopen maanden op veel vlakken hebt geholpen. Jouw adviezen waarmee mijn verzameling zinnen met informatie meer een verhaal (kunnen) worden zullen me zeker in de toekomst helpen om de boodschap van mijn onderzoek beter over te brengen.
Ik heb erg veel plezier in mijn werk maar ik zou het niet kunnen volhouden zonder een intensieve hobby voor de afwisseling. Mijn dank gaat daarom ook uit naar de ruiters en amazones van de Kajützhoeve. Mede dankzij jullie inzet en enthousiasme bestaat onze manege nog steeds en kan ik nog steeds intensief bezig zijn met de paardensport op een manier die ik heel erg leuk vind. Jarna, Onno en Paul, we spreken elkaar soms tijden niet maar als ik jullie tegen kom dan heb ik het altijd gevoel dat we elkaar de dag ervoor nog hebben gesproken. Ook in de afgelopen moeilijke jaren hebben jullie weer bewezen dat jullie er op een of andere manier altijd voor mij zijn als het er echt op aankomt. Heel erg bedankt hiervoor.
promovendi en postdocs van de sectie scheikundige technologie van de universiteit van Groningen die mij in de afgelopen periode gastvrij hebben ontvangen en op weg hebben geholpen. In het bijzonder de medewerkers van de technische dienst (Marcel de Vries en Anne Appeldoorn) en de glasblazerij (Maarten Vervoort) want zonder jullie expertise en hulp had ik de configuratie van polymerisatiereactor niet kunnen optimaliseren gedurende het onderzoek. Arjan Kloekhorst wil ik bedanken voor het maken van de schematische weergave van de polymerisatie reactor en die terug te vinden is in meerdere hoofdstukken. Daarnaast heb ik met veel plezier Sunny Qian begeleidt met haar afstudeeropdracht waarmee ze haar master op de universiteit (cum laude) kon afronden. Gert Alberda van Ekenstein heeft niet alleen mijn kennis met betrekking op de DSC op een nivo hoger gezet maar ook nog alle DSC-metingen voor mij uitgevoerd op de universiteit. I also want to thank the reading committee of my dissertation. Prof. dr. A.A. Broekhuis, prof. dr. K.U. Loos and prof. dr. L. Moscicki, thank you for taking the time and effort to read my thesis. Your recommendations with respect to the technological assessment were used to improve this section considerably. Met de in 2012 toegekende subsidie kwam er ook geld beschikbaar om een Labview programma te laten schrijven waarmee de apparatuur van de laboratoriumopstelling kon worden aangestuurd. Dank zij Carya Automatisering (André Buurman en Paul Rijkers) werd de aansturing en monitoring van de polymerisatieopstelling, na een intensieve samenwerking, een stuk eenvoudiger en had ik vanaf dat moment (hoofdstuk 4) maar 1 softwareprogramma nodig (in plaats van 6).
Een deel van het proefschrift is te herleiden naar kennis en vaardigheden die ik mij voor 2010 eigen heb gemaakt. Ik wil daarom mijn (oud-)collega’s van het Nederlands Instituut voor Koolhydraat Onderzoek (NIKO), Packard Bioscience, Perkin Elmer, DSM Biologics, TNO Quality of life en AVEBE dus ook bedanken voor het creëren van een (over het algemeen) zeer prettige en leerzame werkomgeving. In dit kader wil ik in het bijzonder Harry van Lune nog even noemen. Harry, ik weet niet of het nog voor jou zo 123 herkenbaar is maar mijn manier van werken is nog steeds te herleiden naar ervaring die wij samen hebben opgedaan bij Packard Bioscience.
Eigenlijk zou mijn promotietraject in 2008 beginnen maar dat werd uitgesteld omdat ik mee wou helpen om een groot probleem in de semi-droge crosslink lijn van de extruderfabriek op te lossen. In augustus 2009 nam ik mij stellig voor dat vanaf dat moment alles moest wijken voor een goede start van het promotieonderzoek. Het liep echter totaal anders. Begin september bleek Marjan ernstig ziek te zijn en kwam alles voornamelijk in het teken te staan van zeer ingrijpende en langdurige medische behandelingen en het leren omgaan met de nieuwe situatie. Marjan heeft uiteindelijk altijd begrip op weten te brengen voor mijn situatie en dit is een zeer belangrijke reden waarom ik in deze moeilijke periode niet ben gestopt met het promotieonderzoek. Ik ben heel erg blij dat we nog steeds hele goede vrienden zijn ondanks het feit dat we nu niet meer samenwonen. Pap, Mam, Arjan, we hebben intussen samen al heel wat moeilijkheden doorstaan en daarin hebben wij elkaar altijd geholpen waar het maar even mogelijk was. Voor ons is dat vanzelfsprekend maar intussen is het mij wel duidelijk dat deze onderlinge verhouding lang niet in elk gezin aanwezig is. Pap, ook nog bedankt dat je
138
APPENDICESSummary SamenvattingDankwoordPublications
138
APPENDICESSummary SamenvattingDankwoordPublications
140
Peer-viewed journals
1) H.A. van Doren, K.R. Terpstra, Structure-property relationships in D-glucitol derivatives with two geminal hydrocarbon chains. Part 1. Thermotropic and lyotropic liquid-crystalline behaviour, J. Mater. Chem., 5 (1995) 2153-2160
2) J.M. Pestman, K.R Terpstra, C.A. M Stuart, H.A.van Doren, A. Brisson, R.M. Kellogg, J.B.F.N. Engberts, Non-ionic Bolaamphiphiles and Gemini Surfactants Based on Carbohydrates, Langmuir 13 (1997) 6857-6860
3) K. R. Terpstra, A. J. J. Woortman, J. C. P. Hopman, Yellow dextrins: Evaluating changes in structure and colour during processing, Starch, 62 (9) (2010) 449-457
4) K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified amylopectin potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects, To be published.
5) K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch, To be published.
6) K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics, To be published.
7) K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and amylopectin potato starch, To be published.
8) K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives, To be published.
Conferences1) K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Starch
derivatives as protective colloid in pVAc dispersions, International polymer forum 2011 (HIPF2011; Hangzhou, China)
2) L. Daniel, K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, The use of starch as a protective colloid in emulsion polymerisations of vinyl acetate (IMTCE2014; Kuala lumpur, Malaysia)
3) K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Potato starch stabilized polyvinyl acetate latexes, 2014 IUPAC World Polymer Congress (MACRO2014; Chiang Mai, Thailand)
Patent
1) K.R. Terpstra , et al, Starch stabilized synthetic latexes (Feasibility study).
140
Peer-viewed journals
1) H.A. van Doren, K.R. Terpstra, Structure-property relationships in D-glucitol derivatives with two geminal hydrocarbon chains. Part 1. Thermotropic and lyotropic liquid-crystalline behaviour, J. Mater. Chem., 5 (1995) 2153-2160
2) J.M. Pestman, K.R Terpstra, C.A. M Stuart, H.A.van Doren, A. Brisson, R.M. Kellogg, J.B.F.N. Engberts, Non-ionic Bolaamphiphiles and Gemini Surfactants Based on Carbohydrates, Langmuir 13 (1997) 6857-6860
3) K. R. Terpstra, A. J. J. Woortman, J. C. P. Hopman, Yellow dextrins: Evaluating changes in structure and colour during processing, Starch, 62 (9) (2010) 449-457
4) K.R. Terpstra, F. Picchioni, L. Daniel, G.O.R. Alberda van Ekenstein, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified amylopectin potato starch stabilized polyvinyl acetate latexes: A systematic study on polymerization aspects, To be published.
5) K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, A systematic study on synthesis and properties of polyvinyl acetate latexes stabilized by pyrodextrinated potato starch, To be published.
6) K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Modified waxy potato starch stabilized polyvinyl acetate latexes: Influence of polymerization temperature and initiator concentration on process and product characteristics, To be published.
7) K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Extruded octenyl succinylated starch stabilized polyvinyl acetate latexes: A comparison between regular and amylopectin potato starch, To be published.
8) K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, The use of polyvinyl acetate latexes stabilized by extruded octenyl succinate (waxy) potato starch as wood adhesives, To be published.
Conferences1) K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Starch
derivatives as protective colloid in pVAc dispersions, International polymer forum 2011 (HIPF2011; Hangzhou, China)
2) L. Daniel, K.R. Terpstra, F. Picchioni, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, The use of starch as a protective colloid in emulsion polymerisations of vinyl acetate (IMTCE2014; Kuala lumpur, Malaysia)
3) K.R. Terpstra, F. Picchioni, L. Daniel, A.A.M. Maas, J.C.P. Hopman, H.J. Heeres, Potato starch stabilized polyvinyl acetate latexes, 2014 IUPAC World Polymer Congress (MACRO2014; Chiang Mai, Thailand)
Patent
1) K.R. Terpstra , et al, Starch stabilized synthetic latexes (Feasibility study).